Novel bis-acac-O,O-Ir(III) catalyst for anti-Markovnikov, hydroarylation of olefins operates by arene CH activation

Article abstract of DOI:10.1039/b208680h

A novel, thermally stable, homogeneous Ir catalyst for the anti-Markovnikov, hydroarylation of olefins is shown to operate by arene CH activation via the formation of a bis-acac-O,O phenyl-Ir(III) species.

Full text of DOI:10.1039/b208680h

Novel bis-acac-O,O–Ir(III) catalyst for anti-Markovnikov, hydroarylation
of olefins operates by arene CH activation†
Roy A. Periana,* Xiang Y. Liu and Gaurav Bhalla
University of Southern California, Department of Chemistry, Loker Hydrocarbon Institute, Los Angeles, CA
90089, USA. E-mail: rperiana@usc.com; Fax: +213-821-2656; Tel: +213-821-2035
Received (in Purdue, IN, USA) 5th September 2002, Accepted 24th October 2002
First published as an Advance Article on the web 13th November 2002
A novel, thermally stable, homogeneous Ir catalyst for the
anti-Markovnikov, hydroarylation of olefins is shown to
operate by arene CH activation via the formation of a bis-
acac-O,O phenyl–Ir(III) species.
solubility of 2-H2O and 1 in neat benzene complicated studies).
Attesting to the thermal stability of the bis-acac-O,O–Ir(III
)
motif, refluxing 2-Py with mixtures of benzene and/or neat
acetic acid, hexafluoroacac or 2-methylacetylacetone for sev-
eral hours led only to exchange of the g-C bonded acac ligand
(acac-C3) but no exchange or loss of the bis-acac-O,O bonded
ligands. The reaction mixtures remained homogeneous and only
bis-acac-O,O–Ir(III) species were observed.
Recently we reported the binuclear complex, [Ir(m-acac-
O,O,C3)(acac-O,O)(acac-C3)]2, 1, that catalyzes the only re-
ported example of anti-Markovnikov, hydroarylation of un-
activated olefins by unactivated arenes to produce saturated
alkyl arenes.1 To account for the anti-Markovnikov re-
gioselectivity, it was presumed that the active catalyst was a
homogeneous Ir species that catalyzed reaction by arene C–H
bond activation (rather than e.g., by carbocations intermediates)
but no mechanistic work has been reported to confirm these
possibilities. It is important to understand the basis of operation
of this catalyst as the reactivity is unique and knowledge of the
composition of the active catalyst and reaction mechanism
could provide the basis for developing new classes of catalysts
for regio and stereoselective C–H and C–C bond forming
reactions. In contrast to this system, reported systems that
catalyze reactions between olefins and arene C–H bonds
typically generate unsaturated, rather than saturated products.2
Herein, we report the first evidence that the active catalyst is a
thermally stable, homogeneous, bis-acac-O,O–Ir(III) species
and the reaction proceeds via arene CH activation to generate a
bis-acac, phenyl–Ir(III) intermediate that has been observed and
characterized.
The initial catalysis was observed with the dinuclear Ir
complex 1.1 To examine the possibility that a mononuclear Ir
complex formed from 1 was the active catalyst we sought to
synthesize and examine the reactivity of stable, mononuclear
bis-acac-O,O–Ir(III) complexes. Interestingly, we have found
that the new mononuclear complex,2-H2O, Scheme 1,3 can be
isolated in 45% yield as an initially formed species during the
reported synthesis of 1.4 To examine the possibility that the acac
ligands may be thermally labile at the typical conditions for
catalysis, 150–200 °C for 2 h, we examined the reaction kinetics
and thermal stability of 2-Py, Py = pyridine (the limited
To further determine if the dinuclear Ir species 1 or a
mononuclear Ir species was the active catalyst we compared the
catalytic hydroarylation activity of 1 and the mononuclear
species 2-Py and 2-H2O. As can be seen in Table 1, entries 1–3,
the mononuclear complexes are both active catalysts with
2-H2O being as active as the dinuclear species, 1, while 2-Py is
less active. Importantly, both complexes gave identical linear to
branched (L+B) product ratios compared to 1 in the hydro-
phenylation of propylene. This strongly indicates that the same
active catalytic species (albeit at different concentrations to
account for the differences in rate) is formed from 1, 2-H2O or
2-Py. The lower rate of 2-Py relative to 2-H2O and 1 is
consistent with the better ligating properties of pyridine and the
possibility that the active catalytic species is a mononuclear,
bis-acac-O,O–Ir(III) species formed by dissociation of 1 or
ligand loss from 2-H2O or 2-Pyr. As expected, the addition of
pyridine decreases the rate of catalysis with 1, 2-H2O and
2-Py.5
Consistent with the possibility that the active catalyst is a
mononuclear bis-acac-O,O–Ir(III) species, the calculated cata-
lyst TOF is independent of the amount of added 2-Py (Table 1,
entries 3–5). Assuming equilibrium between a mononuclear and
dinuclear Ir species and that only one of these species is the
active catalyst, this constant TOF would only be expected if
either all the added complex remains mononuclear or all is
converted to a dinuclear species. These results, coupled with the
observation that 2-Py is stable with respect to the formation of
dinuclear species at elevated temperatures in benzene show that
it is unlikely that dinuclear Ir species are the active catalysts.
Supporting this, we have also observed that treatment of 1 with
ethylene at moderate pressures leads to the formation of a
Table 1 Results from the hydrophenylation of propylene catalyzed by bis-
acac-O,O–Ir(III) complexesa
[Complex]
mmol
TOFb
(31024 s21
L+Bc
ratio
Entry Complex
Additive
)
1
2
3
4
5
6
7
8
9
1
—
—
—
—
—
—
—
Acac-H
H2O
5
5
5
10
30
5
5
5
5
110
100
24
22
22
32
130
15
61+39
61+39
61+39
61+39
61+39
61+39
61+39
61+39
61+39
2-H2O
2-Py
2-Py
2-Py
3-Py
3-H2O
3-Py
3-Py
36
Scheme 1 Synthesis of mononuclear bis-acac-O,O–Ir(III) complexes.
a All reactions were carried out at 0.96 MPa of propylene in neat benzene at
180 °C and for 30 min. b TOF = [mols of product]/([mols of added
2-Py].Reaction time). c Mole ratio of linear to branched products.
† Electronic supplementary information (ESI) available: experimental
section. See http://www.rsc.org.suppdata/cc/b2/b208680h/
3000
CHEM. COMMUN., 2002, 3000–3001
This journal is © The Royal Society of Chemistry 2002

View Article Online
mononuclear bis-acac-O,O–Ir(III) species which we speculate is
the ethylene complex, 2-C2H4.
To determine whether the hydroarylation occurred by
benzene C–H activation to generate phenyl–Ir intermediates, we
examined the reaction mixture for phenyl–Ir complexes. NMR
analysis of crude reaction mixtures resulting from reaction of
benzene and propylene, catalyzed with 5 mol% 2-Py, after
removal of volatiles, led to the detection of the phenyl–Ir(III
)
species, Ir(Ph)(acac-O,O)2(pyridine), 3-Py, in ~ 65% yield
based on added 2-Py. The identity of this material has been
confirmed by comparison with the independently synthesized
and fully characterized complex, 3-Py, prepared as shown in
Scheme 1.3 The identical material is also formed when catalysis
is carried out with 2-H2O, followed by addition of pyridine and
isolation.
To confirm that bis-acac-O,O, phenyl–Ir(III) species were
involved in the catalysis, the catalytic activity of 3-Py and the
related complex, 3-H2O (both prepared by independent synthe-
ses, Scheme 1)3 were examined as hydroarylation catalysts. As
can be seen in Table 1, 3-Py (entry 6) was found to be as active
as 2-Py (entry 3) while 3-H2O (entry 7) was found to be slightly
more active than 1 or 2-H2O. Critically, these bis-acac-O,O,
phenyl–Ir(III) complexes also gave the same L+B ratios of
hydroarylation products with propylene strongly indicating that
the same catalytic species are involved starting from 2 or 3. The
efficient reaction of 3-Py seemed to indicate that neither g-C
bonded acac (present in complexes 1 and 2) nor water (present
in 2-H2O) was essential to catalysis. To further confirm this, we
explicitly examined the effect of added free acetylacetone
(acac-H) and water on the catalytic activity of 3-Py in dry
benzene/olefin mixtures. As can be seen in Table 1, (entries 8
and 9) neither added acac-H nor water has any effect on the
reaction rate or product selectivity of 3-Py.6
To provide additional evidence for the intermediacy of a bis-
acac-O,O, phenyl–Ir(III) species in the catalytic, hydrophenyla-
tion of olefins, the stoichiometric reaction of 3-Py and 3-H2O
with propylene in mesitylene,7 saturated with water to provide
a H source as shown in eqn. (1), was examined. This led (based
on added phenyl–Ir complex) to the formation of benzene
(33%) and n-propylbenzene (40%) and isopropylbenzene
(27%). Importantly the ratio of propylbenzene products was in
the same L+B ratio ( ~ 61+39) as observed in the catalytic
hydrophenylation of propylene catalyzed by 1, 2-H2O, 2-Py,
3-H2O or 3-Py. The identity of the Ir products in eqn. (1) were
Fig. 1 Proposed CH activation reaction mechanism for hydroarylation of
olefins by a bis-acac,O,O–Ir(III) species.
invoked for catalysis by 3-Py in dry, neat benzene it is unlikely
that an electrophilic mechanism via cationic 5-coordinate Ir(III
)
species is involved. Instead, we now favor the mechanism
shown in Fig. 1, where arene C–H activation and product
formation occur in the same step from a cis, bis-acac-O,O,
arene, phenethyl Ir(III) species, 4, that generates the product by
either oxidative addition to a C–H bond of the coordinated arene
to generate a seven coordinate Ir( ) intermediate, 5, or by a
V
sigma-bond metathesis transition state. In light of the recent
precedent by Bergman for the involvement of seven coordinate,
Ir(
) intermediates in C–H activation reactions with Ir(III),9 we
V
are biased toward the oxidative addition pathway and are
carrying theoretical calculations to distinguish between the two
possibilities. We have no evidence for the proposed cis-
isomerization but this is reasonable given the requirement for an
olefin insertion step.
Work is continuing on delineating additional details of the
reaction mechanism, understanding the molecular basis for the
uniqueness of the acac-O,O ligands and designing improved
catalysts. A unique aspect of this catalysis is the complete lack
of any olefinic products that could result by b-hydride
elimination from intermediates 4 or 5. As developing strategies
to design catalysts that disfavor such b-hydride eliminations
would be important, we are investigating the synthesis and
chemistry of the phenethyl intermediate, 4, to determine if such
species are formed and why olefins products are not pro-
duced.
not confirmed but it is presumed that a bis-acac-O,O–Ir(III
)
hydroxyl species is formed.
(1)
The results above, along with the initially noted high catalytic
stability and reaction reproducibility1a support the conclusion
that the catalysis proceeds by arene CH activation via a
thermally stable, homogeneous, mononuclear, bis-acac-O,O
aryl–Ir(III) species. While we cannot completely rule out other
Notes and references
1 (a) T. Matsumoto, D. J. Taube, R. A. Periana, H. Taube and H. Yoshida,
J. Am. Chem. Soc., 2000, 122, 7414; (b) T. Matsumoto, R. A. Periana, D.
J. Taube and H. Yoshida, J. Mol. Catal. A, 2002, 180, 1; (c) T. Matsumoto
and H. Yoshida, Catal. Lett., 2001, 72, 107.
2 V. Ritleng, C. Sirlin and M. Pfeffer, Chem. Rev., 2002, 102, 1731.
3 See supporting material for syntheses of new complexes in Scheme 1.
4 M. A. Bennett and T. R. B. Mitchell, Inorg. Chem., 1976, 15, 2936.
5 Pyridine is inert to the hydroarylation reaction. Exchange can be observed
between pyridine-d5 and 3-Py.
6 As a result of more detailed studies, added water, rather than accelerating
catalysis by 1 as reported earlier,1a has now been found to have little
effect on the rate of catalysis by 1 or 2. This result is consistent with the
proposed CH activation reaction mechanism.
7 Mesitylene is an inert solvent that supports the hydroarylation catalysis.
The arene CH bonds of mesitylene are not activated.
possibilities, such as Ir( ) species or slippage of the spectator
I
2-
1
h acac-O,O ligands to an h -acac ligand that is O or C bound,
we believe that formation of such species are not consistent with
the complete lack of exchange of the bis-acac-O,O ligands with
free acac-H at elevated temperatures or in the presence of acids.
Additionally, the relatively low level of inhibition by olefins1
would not be expected with an Ir(
I) species as the active
species.8
Interestingly, if the catalysis is based on Ir(III) rather than an
) species, an important question is how the arene C–H bond
is activated. Our initial speculation1b of an electrophilic
substitution mechanism by a 5-coordinate, cationic Ir(III
Ir(
I
)
8 M. Kanzelberger, B. Singh, M. Czerw, K. Krogh-Jespersen and A. S.
Goldman, J. Am. Chem. Soc., 2000, 122, 11017.
9 S. R. Klei, ST. D. Tilley and R. G. Bergman, J. Am. Chem. Soc., 2000,
122, 1816; P. J. Alaimo and R. G. Bergman, Organometallics, 1999, 18,
2707.
species formed by dissociation of the anionic, g-C bonded acac
(acac-C3) was based on early results that implicated that the g-C
bonded acac-C3 ligand of 1 was essential to catalytic activity.
However, as no such plausible anionic leaving group can be
CHEM. COMMUN., 2002, 3000–3001
3001