Transition-metal complexes with sterically demanding ligands: Facile thermal intermolecular C-H bond activation in a square-planar Ir1 complex

Article abstract of DOI:10.1002/anie.200219908

Teaching an old dog a new trick: A novel square-planar iridium(I) system undergoes facile thermal intermolecular C-H activation process at ambient temperature (see scheme). The observed unusual low activation barrier is analyzed and traced by DFT methods which provide guidelines for the design of novel systems, which would allow the activation of methane under mild conditions.

Full text of DOI:10.1002/anie.200219908

Communications
C-H Bond Activation
Transition-Metal Complexes with Sterically
Demanding Ligands: Facile Thermal
ꢀ
Intermolecular C H Bond Activation in a
Square-Planar IrI Complex**
Stefan Nückel and Peter Burger*
Dedicated to Professor Robert G. Bergman
on the occasion of his 60th birthday
ꢀ
The intermolecular activation of C H bonds is a highly
desirable process[1–8] and is part of many catalytic systems,[9–11]
for example, the dehydrogenation of alkanes to alkenes.[12–14]
Detailed studies by various groups clearly established that the
ꢀ
C H activation process is under kinetic rather than thermo-
dynamic control in most cases.[12,13] Thus, highly reactive
species, such as photochemically generated short-lived tran-
sients, for example, [Cp*Ir(PMe3)] (Cp* = C5Me5), are often
required,[3,15] while thermally initiated C H activation proc-
ꢀ
esses under mild conditions are less common.[2,4,16] Herein, we
describe a novel square-planar iridium(i) system 1, which
ꢀ
undergoes facile thermal intermolecular C H activation in
benzene at ambient temperature [Eq. (1)]. The observed
N
N
N
N
C6H6, 40°C, 2h
- CH4
N
N
Ir
Me
Ir
Ph
(1)
1
2
quantitative
unusual low barrier is analyzed and traced by DFT methods
providing guidelines for the design of novel systems, which
would allow the activation of methane under mild conditions.
We have recently reported the synthesis of the iridium
complex 1 and its rhodium analogue.[1] In the course of this
ꢀ
study, we noticed facile intermolecular C H activation in
benzene leading quantitatively to the phenyl complex 2 and
methane [Eq. (1)]. In the analogous reaction of 1 with C6D6
formation of [D5]phenyl analogue of 2 and [D1]methyl was
observed. Both 1 and 2 could be characterized unambiguously
[*] Dr. P. Burger, Dr. S. Nückel
Anorg.-chem. Institut
Universität Zürich
Winterthurerstrasse 190, 8057 Zürich (Switzerland)
Fax: (+41)1-635-6802
E-mail: chburger@aci.unizh.ch
[**] Part IV. We would like to thank the Swiss National Science
Foundation for financial support. We are indebted to Prof. Heinz
Berke for his generous support. For Part III see ref. [1].
1632
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200219908
Angew. Chem. Int. Ed. 2003, 42, 1632 – 1636

Angewandte
Chemie
as square-planar complexes by their X-ray crystal structures
(Figure 1).[17] Compound 2 was also obtained by independent
synthesis from the corresponding methoxy complex[1] and
phenylmagnesium chloride in 38% yield.
notion of the two-step mechanism. Preliminary experimental
evidence for an IrIII intermediate was provided by the reaction
of 1 with H2, which led to a tentatively assigned, trihydride IrIII
complex and methane.[29] The two-step view was partially
supported by the absence of a transition state for the s-bond
metathesis pathway in our DFT calculations. Furthermore, we
ꢀ
considered C H bond activation at a three-coordinate metal
ꢀ
center, with C H activation in benzene by [Ir(PiPr3)2Cl] being
a well-established example.[30] Related results were recently
reported for an iridium cycloalkane-dehydrogenation catalyst
bearing a terdentate cyclometalated bisphosphane donor
(PCP pincer ligand).[13] Considering the rigid chelating ligand
in 1, the cleavage of one of the nitrogen donor arms from the
metal center was deemed to be inconsistent with the observed
low activation energy. This idea was supported by the DFT
ꢀ
calculations, which revealed an Ir Ndiimine bond-dissociation
enthalpy of 35 kcalmolꢀ1.
These results suggested that an alternative mechanistic
scenario was operative; we have therefore studied the two-
step mechanism starting from the square-planar complex 1.
For this pathway, we were able to identify the required
Figure 1. ORTEP plots (thermal elipsoids set at 50% probability) of
complexes 1 (top) and 2 (bottom).
III
ꢀ
transition states TSOA and TSRE and the C H oxidative Ir
intermediate IOA (OA = oxidative addition, RE = reductive
elimination) as stationary points[31] (Figure 2).
ꢀ
Monitoring the C H activation process shown in Equa-
tion (1) by H NMR spectroscopy indicated a clean conver-
The calculated barrier of 22 kcalmolꢀ1 leading from 1 via
the square-pyramidal transition state TSOA to the octahedral
H5C6–H (with benzene) oxidative-addition product IOA is in
excellent agreement with the observed experimental value. It
is noteworthy that the IrIII intermediate, IOA, is energetically
“uphill” (+ 10 kcalmolꢀ1) from 1 and benzene. This readily
explains the calculated low barrier of 12 kcalmolꢀ1 for the
reductive elimination of CH4 passing over TSRE to give the
1
sion from 1!2 without detectable traces of intermediates. In
neat [D6]benzene under pseudo first-order conditions, a first-
order decay of the starting material 1 with a free enthalpy of
activation, DG313 = 23 kcalmolꢀ1 was measured.
ꢀ
Overall, the outcome of the C H activation reaction
resembles a s-bond metathesis process, which is the antici-
pated mechanism for d0 early-transition-metal and lanthanide
row complexes.[8,18,19] For dn-configured complexes with n ꢁ 2
energetically favored final products
2 and methane
(ꢀ4 kcalmolꢀ1). For the corresponding Rh system, 1Rh, we
calculated a significantly higher barrier of 27 kcalmolꢀ1,
which is consistent with the observed reluctance of 1Rh to
ꢀ
there are other mechanistic alternatives, for example, a C H
ꢀ
oxidative-addition step, followed by C H reductive elimina-
tion.[4,20,21] Investigations of late-transition-metal systems of
which the cationic IrIII compounds [Cp*Ir(PMe3)Me]+Xꢀ are
prominent examples,[16,22] indicate that the two-step pathway
is followed for these systems. However, it is extraordinary
difficult to distinguish between these two mechanistic scenar-
undergo C H activation in benzene even at elevated temper-
atures.
ꢀ
ꢀ
The results of the DFT calculations for C H activation in
methane by complex 1 are also included in Figure 2. In
accordance with previous studies,[2,3] the barrier is higher (by
ꢀ1
ꢀ
ꢀ
ios in the absence of experimental evidence for a C H
+ 6 kcalmol ) than for the C H activation process in
oxidative intermediate.[2,4,20,21,23] In the aforementioned IrIII
system, support for the two-step pathway was provided by
benzene. For the analogous less crowded, 1H4 (Figure 2)
however, we derived a substantially smaller barrier of just
DFT calculations prior to the recent isolation of a related IrV
19 kcalmolꢀ1 for the C H oxidative-addition step in methane
ꢀ
[20,21]
(Figure 2). Using a value of 9 kcalmolꢀ1 for the TDS° term
(DS° = + 30 calmolꢀ1 Kꢀ1) at RT, a free enthalpy of activation
of DG° ꢂ 28 kcalmolꢀ1 is estimated. This result suggested that
methane could be activated under relatively mild thermal
conditions in the 1H4 system.
ꢀ
ꢀ
Si H and C H oxidative-addition product.
ꢀ
To elucidate mechanistic details of the C H activation
system in the square-planar complex 1, we also resorted to
DFT methods. Since steric factors were deemed to play a
ꢀ
crucial role in the C H activation process, DFT calculations
ꢀ
were performed on the full system. The efficient parallel RI-
DFT implementation[24,25] (BP-86 functional; basis: TZVP,
ECP-60-MWB (Ir), SVP (C,H,N)) of the Turbomole program
package[26] allowed the use of a pure quantum-mechanical
approach rather than the potentially less reliable quantum
mechanical/molecular modeling methods,[27,28] which have
been applied by other groups for systems of related size.
Although we did not intend to prejudge the mechanism,
after considering the d8-configuration of 1 we preferred the
To understand the unprecedented low barriers for the C
H oxidative-addition process in the square-planar systems 1
[32]
and 1H4
,
we performed an intrinsic reaction coordinate
[33]
(IRC) calculation
starting from the square-pyramidal
(Figure 2) transition state TSOA for the methane oxidative-
addition step in 1H4. In the forward direction, this led to the
anticipated octahedral IrIII product IH4,CH4. Following the
backward reaction mode, the anticipated lengthening of the
ꢀ
ꢀ
ꢀ
Ir H and Ir C and shortening of the C H bond was
Angew. Chem. Int. Ed. 2003, 42, 1632 – 1636
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1633

Communications
Me
N
H
R
N
Ir
N
N
Me
H
"square pyramidal"
N
Ir
∆Erel
/
N
kcal mol-1
TSOA
Me
"square pyramidal"
Ph
R = Me
+25(+19)
+22(+16)
N
TSRE
N
Ir
H
20
N
R = Ph
+22(+15)
octahedral
R
IOA
+16(+7)
+10(+8)
R = Me
R = Ph
10
N
N
Ir
Me
N
N
N
Ir
Ph
N
0
+ RH: R = Me, Ph
+ CH4
-4(-5)
N
values in parenthesis for
N
N
Ir
Me
1H4
ꢀ
Figure 2. Energy diagram and proposed intermediates and transition states for the C H activation process shown in Equation (1) based on DFT
theory.
observed. More importantly, substantial opening of the
pyridyl-Ir-Me angle from approximately 908 in the transition
state was noticeable, and significant bending was still
observed after methane was essentially fully released
an even more shallow potential for the bending process, with
the Walsh diagrams of 1model and 1H4 being essentially
superimposable (Figure 3). Since identical kinetic and ther-
N
ꢀ
modynamic data for the methane C H oxidative activation
ꢀ
(d(Ir···H CH3) = 2.5 Š). This result clearly indicated bending
process were also derived for 1model and 1H4, we anticipate that
the lower barriers in the N-phenyl substituted complex 1H4
(compared to 1) can be attributed to the more shallow
potential of the bending process, which is a consequence of
the reduced steric crowding. This situation is in contrast to the
Walsh diagram obtained for this process in the tridentate
phosphorus donor complex, [{k3-MePCH2CH2PMe2)2}IrMe].
For the latter, a significantly steeper potential for the bending
process is observed and therefore a substantially larger
contribution to the activation energy is anticipated. Before
the origin of the difference between the N and P ligands will
be addressed, the electronic structure of the bent species will
be discussed.
of the methyl group away from the chelating-ligand–metal
ꢀ
plane in the Ir complex as a crucial part of the C H activation
process and prompted us to calculate a Walsh diagram for this
out-of-plane bending in complex 1 (Figure 3).
The small energy required for out-of-plane bending of the
methyl group in complex 1 is noteworthy, approximately
12 kcalmolꢀ1 for a = 708. For the unsubstituted model com-
¼
pound [{2,6-[C(CH3) NH]2pyridine}IrMe] (1model) we noticed
As previously analyzed in EHT and DFT studies, an out-
of-plane bending distortion of one ligand in a d8-configured
square-planar ML4 complex leads to a C2v-symmetrical ML4
species which is isolobal to the methylene carbene unit,
CH2.[34,35] The [Fe(CO)4] singlet state is a prominent repre-
sentative of this type of compound. In the bent complex 1, this
is also nicely reflected in the frontier orbitals (a = 908) which
have the required p(HOMO) and s(LUMO) symmetry
ꢀ
(Figure 4). Small C H activation barriers would therefore
be expected for this type of species. It should be noted that for
angles a < 908, a residual two-orbital four-electron repulsive
term between a filled metal d orbital and the occupied
s CH orbital raises the barrier.[35] This term vanishes when a
bending angle of 908 is attained, as observed in the transition
state TSOA of complex 1.
Figure 3. Walsh diagram for out-of-plane bending of [{k3-MeP(CH2CH2P-
!
¼
&
*
^
Me2)2}IrMe] ( ), complex 1 ( ), 1model ( ), 1H4 ( ), and [{2,6-[C(CF3) NH]2-
~
pyridine}IrMe] ( ).
1634
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 1632 – 1636

Angewandte
Chemie
ꢀ
Keywords: C H bond activation · density functional theory ·
iridium · N ligands
.
[1] S. Nückel, P. Burger, Organometallics 2001, 20, 4345 – 4359.
[2] B. A. Arndtsen, R. G. Bergman, T. A. Mobley, T. H. Peterson,
Acc. Chem. Res. 1995, 28, 154 – 162, and references therein.
[3] R. G. Bergman, J. Organomet. Chem. 1990, 400, 273 – 282, and
references therein.
Figure 4. Frontier orbitals of complex 1.
[4] R. H. Crabtree, J. Chem. Soc. Dalton Trans. 2001, 2437 – 2450.
[5] C. L. Hill, Activation and Functionalization of Alkanes, Wiley-
Interscience, New York, 1989, and references therein.
[6] A. E. Shilov, G. B. Shul'pin, Chem. Rev. 1997, 97, 2879 – 2932,
and references therein.
[7] G. A. Luinstra, J. A. Labinger, J. E. Bercaw, J. Am. Chem. Soc.
1993, 115, 3004 – 3005.
[8] P. L. Watson, J. Am. Chem. Soc. 1983, 105, 6491 – 6493.
[9] A. Sen, Acc. Chem. Res. 1998, 31, 550 – 557.
Out-of-plane bending potentials in d8-configured square-
planar ML4 complexes have been discussed in the context of
oxidative addition of H2.[36,37] The major contribution to this
unfavorable process arises from a two-orbital-four-electron
repulsive interaction between the occupied metal-based dxz
orbital and the methyl s orbitals upon bending, which is
nonbonding in the undistorted square-planar ML4 complex.
As shown in Figure 5 this repulsive interaction is significantly
lowered by mixing of a diimine, pyridine-ligand-based p-
acceptor orbital in the iridium methyl complex 1.
[10] R. A. Periana, D. J. Taube, S. Gamble, H. Taube, T. Satoh, H.
Fujii, Science 1998, 280, 560 – 564.
[11] R. A. Periana, D. J. Taube, E. R. Evitt, D. G. Loffler, P. R.
Wentrcek, G. Voss, T. Masuda, Science 1993, 259, 340 – 343.
[12] W. W. Xu, G. P. Rosini, M. Gupta, C. M. Jensen, W. C. Kaska, K.
Krogh-Jespersen, A. S. Goldman, Chem. Commun. 1997, 2273 –
2274.
[13] M. Kanzelberger, B. Singh, M. Czerw, K. Krogh-Jespersen, A. S.
Goldman, J. Am. Chem. Soc. 2000, 122, 11017 – 11018.
[14] R. H. Crabtree, The Organometallic Chemistry of the Transition
Metals, Wiley-Interscience, New York, 2001.
[15] S. E. Bromberg, H. Yang, M. C. Asplund, T. Lian, B. K.
McNamara, K. T. Kotz, J. S. Yeston, M. Wilkens, H. Frei, R. G.
Bergman, C. B. Harris, Science 1997, 278, 260 – 263.
Figure 5. Lowering of the repulsive interaction between the metal dxz
and the methyl s orbitals by mixing of a p-acceptor orbital in complex
1.
[16] B. A. Arndtsen, R. G. Bergman, Science 1995, 270, 1970 – 1973.
[17] Data collection: Stoe IPDS Image Plate System, absorption
corrections were carried out numerically. Structure solution:
Direct methods using the SHELXS-97 program package; refine-
ment against F2 with SHELXL-97.[38] All non-hydrogen atoms
were refined anisotropically; the position of the hydrogen atoms
were calculated in idealized positions and refined using a riding
model. Complex 1: Green needle (0.37 ” 0.11 ” 0.09 mm). Ortho-
rhombic cell, space group P212121 (No. 19), Z = 4, a = 12.914(1),
The major stabilizing contribution stems from back
¼
donation into the -C Ndiimine unit. It has to be emphasized
that this interaction is nonbonding if the ligands are located in
the nodal plane of the dxz orbital. As can be readily seen from
the Ndiimine-Ir-Ndiimine angle of 1578 in the X-ray crystal
structure of 1, the diimine Ndonor atoms are significantly
“bent back” from the Ir-Me unit. This arrangement is a
consequence of the ligand constraints, that is, the “too-short”
b = 13.021(1), c = 14.0217(8) Š, V= 2357.8(3) Š3, 1calcd
=
1.625 gcmꢀ3. 18190 reflections collected, thereof 4360 independ-
ent; 4360 reflections used for structure solution and refinement.
The refinement converged at final R values R(I > 2s(I)): R1 =
0.0150 and wR2 = 0.0301 for 278 parameters; GOF 1.01, residual
electron density: 0.51, ꢀ0.5 eŠꢀ3 (FlackX-Parameter ꢀ0.030(6)
and 1.05(2) for the inverted structure). Complex 2: The crystals
were of low quality (presumably twinned); repeated attempts
under varied conditions did not provide better single crystals.
Brown rectangular parallelepiped (0.3 ” 0.3 ” 0.2 mm); triclinic
ꢀ
C N bonds, which enable strong backdonation. Further
evidence for this view was provided by the Walsh diagram of
¼
the CF3 substituted diimine model compound, [{2,6-[C(CF3)
NH]2pyridine}IrMe] (Figure 3) bearing a stronger p-acceptor
ligand. This situatiuon leads to a preference for the bent over
the square-planar structure with a calculated minimum at a =
1548. Finally, this rationale also serves to explain the larger
barrier for the bending process in the aforementioned
triphosphorus donor complex. Inspection of the DFT opti-
mized structure revealed that the s* phosphorus p-acceptor
orbitals are located in close proximity to the dxz nodal plane
and thus, significantly less stabilization through backdonation
can be achieved.
ꢀ
cell, space group P1 (No. 2), Z = 2 with two independent
molecules per unit cell. a = 10.020(1), b = 14.688(1), c =
18.741(2) Š, a = 85.13(1)8, b = 89.65(1)8, g = 72.31(1)8. V=
2617.8(2) Š3, 1calcd = 1.628 gcmꢀ3. 9272 independent reflections
used for structure solution and refinement. Final R-values (635
parameters) R(I > 2s(I)): R1 = 0.1156 and wR2 = 0.3235.
CCDC-191048 (1) CCDC-203880 (2) contains the supplemen-
tary crystallographic data for this paper. These data can be
obtained free of charge via www.ccdc.cam.ac.uk/conts/retrie-
ving.html (or from the Cambridge Crystallographic Data Centre,
12 Union Road, Cambridge CB21EZ, UK; fax: (+ 44)1223-336-
033; or deposit@ccdc.cam.ac.uk).
ꢀ
We have presented a novel thermal C H activation
system and have traced the low activation energies by DFT
methods, sterically less crowded complexes are predicted to
ꢀ
display even lower C H activation barriers.
[18] P. L. Watson, G. W. Parshall, Acc. Chem. Res. 1985, 18, 51 – 56.
[19] P. L. Watson, J. Chem. Soc. Chem. Commun. 1983, 276 – 277.
[20] S. R. Klei, T. D. Tilley, R. G. Bergman, J. Am. Chem. Soc. 2000,
122, 1816 – 1817.
Received: August 8, 2002
Revised: December 30, 2002 [Z19908]
Angew. Chem. Int. Ed. 2003, 42, 1632 – 1636
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1635

Communications
[21] S. Q. Niu, M. B. Hall, J. Am. Chem. Soc. 1998, 120, 6169 – 6170.
[22] P. Burger, R. G. Bergman, J. Am. Chem. Soc. 1993, 115, 10462 –
10463.
[23] T. M. Gilbert, I. Hristov, T. Ziegler, Organometallics 2001, 20,
1183 – 1189.
[24] K. Eichkorn, O. Treutler, H. ꢀhm, M. Hꢁser, R. Ahlrichs, Chem.
Phys. Lett. 1995, 242, 652 – 660.
[25] K. Eichkorn, O. Treutler, H. ꢀhm, M. Hꢁser, R. Ahlrichs, Chem.
Phys. Lett. 1995, 240, 283 – 289.
[26] R. Ahlrichs, M. Bꢁr, M. Hꢁser, H. Horn, C. Kꢂlmel, Chem. Phys.
Lett. 1989, 162, 165 – 169.
[27] S. Dapprich, I. Komaromi, K. S. Byun, K. Morokuma, M. J.
Frisch, J. Mol. Struct. 1999, 462, 1 – 21.
[28] J. Gao in Reviews in Computational Chemistry, Vol. 7 (Eds.:
D. B. Boyd, K. B. Lipkowitz), Wiley, New York, 1996.
[29] S. Nückel, Dissertation, Universitꢁt Zürich 2002.
[30] H. Werner, A. Hꢂhn, M. Dziallas, Angew. Chem. 1986, 98, 1112 –
1114; Angew. Chem. Int. Ed. Engl. 1986, 25, 1090 – 1092.
[31] All stationary points were characterized by the calculation of
second derivatives through the absence (minima) or one
imaginary frequency (transition states). Transition-state optimi-
zations were performed with John Baker's OPTIMIZE program,
available from Parallel Quantum Solutions, Fayetteville, AR,
USA, interfaced with Turbomole.
[32] R. T. Price, R. A. Andersen, E. L. Muetterties, J. Organomet.
Chem. 1989, 376, 407 – 417.
[33] M. Sierka, J. Sauer, J. Chem. Phys. 2000, 112, 6983 – 6996.
[34] J. Y. Saillard, R. Hoffmann, J. Am. Chem. Soc. 1984, 106, 2006 –
2026.
[35] T. Ziegler, V. Tschinke, L. Y. Fan, A. D. Becke, J. Am. Chem.
Soc. 1989, 111, 9177 – 9185.
[36] A. Dedieu, A. Strich, Inorg. Chem. 1979, 18, 2940 – 2943.
[37] S. A. Macgregor, O. Eisenstein, M. K. Whittlesey, R. N. Perutz, J.
Chem. Soc. Dalton Trans. 1998, 291 – 300.
[38] G. M. Sheldrick, SHELXL-97, Programs for Crystal Structure
Solution and Refinement, Universitꢁt Gꢂttingen (Germany)
1997.
1636
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 1632 – 1636