C-H bond activation by rhodium(I) hydroxide and phenoxide complexes

Article abstract of DOI:10.1002/anie.200700270

(Chemical Equation Presented) Rhodium(I) complexes of the form [(PNP)Rh(OR)] (R = H, CH₂CF₃, C₆H₅; PNP = 2,6-bis[(di-tert-butylphosphino)methyl]pyridine) have been prepared. Upon thermolysis in [D₆]be

Full text of DOI:10.1002/anie.200700270

Communications
DOI: 10.1002/anie.200700270
À
C H Bond Activation
À
C H Bond Activation by Rhodium(I) Hydroxide and Phenoxide
Complexes**
Susan M. Kloek, D. Michael Heinekey,* and Karen I. Goldberg*
Catalysts for selective hydrocarbon oxidation are crucial to
more efficient utilization of our hydrocarbon resources. In
recent years, a substantial number of metal complexes
weakly; in the presence of nitrogen, an equilibrium between 1
and the monomeric dinitrogen complex [(PNP)Rh(N2)]OTf
(2) was observed. Complex 2 has been characterized by NMR
and IR spectroscopy, and elemental analysis.
Despite the lability of the triflate ligand, the correspond-
ing RhI alkyl and alkoxide complexes cannot be easily
prepared by salt metathesis; the CH2 groups on the PNP
ligand of 2 are deprotonated by strong bases (MeLi, KOtBu),
generating the neutral dinitrogen complex [(PNÀP)Rh(N2)]
(3) [Eq. (1)].
À
capable of the selective activation of C H bonds under mild
conditions have been discovered.[1] In spite of these advances,
À
practical catalysts for C H bond functionalization remain
À
elusive, owing in part to the fact that the C H bond activation
reaction is often inhibited by water or by the alcohol product
eliminated from the metal complex, particularly for low-
valent late-metal complexes.[1a,2] In this context, a desirable
À
reaction would be the addition of a C H bond across a metal
alkoxide or hydroxide complex, releasing an alcohol and
generating a new metal–carbon bond in one step. In 2005, the
first examples of this reaction class were reported by the
À
groups of Periana and Gunnoe, who documented the C H
bond activation of arenes by iridium(III) methoxide and
ruthenium(II) hydroxide complexes, respectively.[3] These
reactions begin with saturated 18-electron metal complexes,
and are proposed to proceed by ancillary ligand dissociation
and subsequent s-bond metathesis for the arene activation
step.[4] In this contribution, we report arene activation by
rhodium(I) hydroxide, trifluoroethoxide, and aryloxide com-
plexes bearing a PNP pincer ligand (PNP = 2,6-bis[(di-tert-
butylphosphino)methyl]pyridine). A similar s-bond meta-
Complex 3 was characterized by NMR spectroscopy and
elemental analysis, and displays a characteristic AB pattern of
doublets in the 31P NMR spectrum, indicating inequivalent
À
phosphorus atoms. The IR spectrum of 3 shows an N N
stretch at 2122 cmÀ1, slightly shifted from the N N stretch of 2
À
thesis pathway for the C H bond activation may also be
(2153 cmÀ1), thus consistent with a greater degree of back-
bonding to the nitrogen ligand in the neutral compound 3.
Similar deprotonation reactions of the methylene positions in
pincer ligands on Pd, Pt, Ir, and Ru centers have previously
been reported.[5,6]
À
involved in these reactions. However, in contrast to the
previously reported systems, the open site for arene inter-
action is readily available in these rhodium(I) 16-electron
complexes and ancillary ligand dissociation is not required.
The results reported herein not only increase the generality of
Use of the milder alkylating reagent ZnMe2 avoids the
problem of deprotonation of the ligand backbone; the
rhodium(I) methyl complex [(PNP)Rh(CH3)] (4) was effi-
ciently prepared by the reaction of 1 or 2 with ZnMe2. The
initial product of this reaction was a rhodium–zinc adduct,
formulated as [(PNP)Rh(CH3)Zn(CH3)]OTf based on NMR
spectroscopy and elemental analysis. Addition of one equi-
valent of 2,2’-bipyridyl (bipy) to a dark yellow solution of the
rhodium–zinc adduct resulted in an immediate color change
to dark pink and release of 4, as shown in Scheme 1. Complex
4 was isolated by extraction with pentane.[7]
À
the novel mode of C H bond activation but also provide new
opportunities for the development of more active systems.
The rhodium(I) triflate complex [(PNP)Rh(OTf)] (1) was
prepared by the reaction of [{Rh(coe)2Cl}2] (coe = cyclooc-
tene) with silver triflate, filtration, and addition of the PNP
ligand. This complex was characterized by NMR spectroscopy
and elemental analysis. The triflate ligand in 1 is bound quite
[*] S. M. Kloek, Prof. D. M. Heinekey, Prof. K. I. Goldberg
Department of Chemistry
University of Washington
Box 351700, Seattle, WA 98195 (USA)
Fax: (+1)206-685-8665
E-mail: heinekey@chem.washington.edu
goldberg@chem.washington.edu
The isolation of [(PNP)Rh(CH3)] (4) provided a viable
synthetic route to the previously unavailable hydroxide and
alkoxide complexes. Reaction of 4 with water or the
appropriate alcohol in THF at room temperature generated
the rhodium(I) hydroxide, trifluoroethoxide, and phenoxide
complexes [(PNP)Rh(OR)] (R = H (5), CH2CF3 (6), C6H5
(7); Scheme 1). Complexes 5, 6, and 7 were characterized by
NMR spectroscopy and elemental analysis.[7]
[**] This work was supported by the NSF as part of the Center for the
Activation and Transformation of Strong Bonds (CATSB), Grant No.
0434568. We thank W. D. Jones and J. Kovach for helpful
discussions. The crystal structure of 4 was solved by W. Kaminsky.
The thermal reactivity of the {(PNP)Rh} complexes 4–7
was investigated. When 4 was heated (808C) in [D6]benzene,
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4736
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4736 –4738

Angewandte
Chemie
alcohol product of the reaction. The protons at these positions
have already been observed to be acidic (see above). It is of
interest that for the analogous [(PNP)IrPh] complex, deuter-
ation of the methylene groups on the PNP ligand with the
addition of D2 was proposed to result from the proton transfer
from the methylene position to the metal center with
formation of IrIII and subsequent D2 activation at the IrIII
center.[6] A related mechanism for the Rh system cannot be
III
À
excluded, although C D activation of [D6]benzene by Rh
seems unlikely.
Unlike the hydroxide and trifluoroethoxide complexes,
phenoxide complex 7 was stable to prolonged heating in
[D6]benzene at 1008C. No conversion into [D5]-8 was
observed, and reaction of a benzene solution of 8 with
2.5 equivalents of phenol resulted in quantitative conversion
into 7, suggesting that phenoxide complex 7 is the thermo-
dynamically preferred product. However, complex 7 proved
to be an effective catalyst for H/D exchange between D2O
and benzene. A solution of 7, D2O, and benzene was heated at
1008C under Ar and the intensity of the residual [D]benzene
signal was monitored by 2H NMR (integration against a
[D8]THF internal standard). The [D1]benzene signal was
found to increase linearly with time, so that approximately
15 turnovers were observed after 100 h (where a turnover is
defined as mol 2H incorporated per mol catalyst). The catalyst
showed no decomposition under these conditions. This
reaction is uncommon for transition metals; few examples
of the homogeneously catalyzed H/D exchange between D2O
and organic molecules have been reported.[3,8]
As strong acids are known to react with arenes through
electrophilic aromatic substitution pathways,[9] control experi-
ments were performed to confirm that the observed H/D
exchange was in fact metal-catalyzed. The possibility that a
small amount of phenol could be produced from 7 under the
reaction conditions and might facilitate the exchange process
was considered. However, when phenol was dissolved in a
mixture of [D8]toluene and H2O and heated at 1008C, no
increase in the residual [D7]toluene signals was detected, even
after 360 h (as determined by integration of the 1H NMR
spectrum against an internal standard). A similar reaction was
conducted with complex 7 in [D8]toluene containing H2O
[Eq. (2)].
Scheme 1. Synthesis of complexes 4–7.
the deuterated phenyl complex [(PNP)Rh(C6D5)] ([D5]-8)
was produced in 95% yield. Complex 8 was independently
prepared from 1 and ZnPh2, and characterized by NMR
spectroscopy and elemental analysis.[7] Remarkably, the other
product of the thermolysis reaction was not CH3D but rather
CH4. This result led us to consider the possibility of
contamination of the reaction by adventitious water. Con-
sistent with this proposal, we found that simply dissolving 4 in
[D6]benzene saturated with H2O (approximately 1–2 equiv) at
room temperature resulted in the formation of both the
hydroxide complex 5 (21%) and the phenyl complex [D5]-8
(73%). In the absence of water, no reaction occurred between
complex 4 and [D6]benzene at room temperature. Thus, water
À
appears to play an important role in the C H activation
reaction involving the RhI methyl complex, suggesting the RhI
hydroxide 5 might be an active species for benzene activation.
Heating
a sample of isolated [(PNP)Rh(OH)] in
[D6]benzene at 608C for 95 h resulted in formation of the
deuterated phenyl complex [(PNP)Rh(C6D5)] ([D5]-8) in
60% yield (Scheme 1). An intractable yellow solid precipitate
was also formed during the reaction; the identity of this
material remains unclear. At 258C, no reaction occurred
between 8 and 1.6 equivalents of H2O in [D6]benzene
solution. Similarly, heating a solution of trifluoroethoxide
complex 6 in [D6]benzene at 1008C for 158 h resulted in
partial conversion into the deuterated phenyl complex [D5]-8
(40%), with unreacted 6 also present (40%). This reaction
was accompanied by some decomposition in the form of a
solid precipitate. Prolonged reaction led to increased decom-
position with no increase in the yield of 8.
In the reactions of 5 and 6 in [D6]benzene, partial
deuterium incorporation (40–60%) into the methylene
bridges of the PNP ligand on the product [D5]-8 was observed.
Efficient deuterium exchange into the methylene positions of
8 was subsequently demonstrated through a control experi-
ment in which reaction of the phenyl complex 8 with D2O
(10 equiv) in C6H6 resulted in deuterium incorporation
(approximately 75%) into the methylene groups over the
course of 24 h at room temperature, as determined by 2H and
1H NMR spectroscopy. Thus, the deuterium incorporation
into the methylene positions can be most directly explained
by proton/deuterium exchange with the deuterated water or
After 335 h at 1008C, approximately 47 turnovers were
observed, with
a selectivity (statistically corrected) of
1.2:1:0:0 meta/para/ortho/CH3. This selectivity is consistent
with a metal-catalyzed reaction, as opposed to an electro-
philic aromatic substitution mechanism, in which ortho and
para activation dominate.[9]
Angew. Chem. Int. Ed. 2007, 46, 4736 –4738ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4737

Communications
In contrast to the reactivity of 5 and 6, but similar to that
observed for 7, when the RhI triflate 1 was heated in
[D6]benzene at 1008C, no conversion into 8 was observed.
However, complex 1 was found to be a much less effective
catalyst for the exchange between benzene and D2O than 7,
resulting in approximately 1.5 turnovers after 100 h at 1008C.
Thus, the basicity of the phenoxide ligand of complex 7
appears to play an important role in the observed H/D
exchange reactions. However, the lone electron pair appears
to also be needed to mediate the s-bond metathesis. The
methyl complex 4 did not react in dry [D6]benzene at room
temperature, with conversion into 8 occurring only in the
presence of H2O, presumably by formation and reaction of
the RhI hydroxide 5 (see above).
[1] a) R. A. Periana, G. Bhalla, W. J. Tenn III, K. J. H. Young, X. Y.
Liu, O. Mironov, C. J. Jones, V. R. Ziatdinov, J. Mol. Catal. A 2004,
220, 7; b) A. S. Goldman, K. I. Goldberg, Activation and Func-
À
tionalization of C H Bonds, Oxford University Press, Washing-
ton, DC, 2004, pp. 1 – 43; c) J. A. Labinger, J. E. Bercaw, Nature
2002, 417, 507; d) R. H. Crabtree, J. Chem. Soc. Dalton Trans.
2001, 2437; e) W. D. Jones, Science 2000, 287, 1942.
[2] a) J. S. Owen, J. A. Labinger, J. E. Bercaw, J. Am. Chem. Soc.
2006, 128, 2005; b) H. A. Zhong, J. A. Labinger, J. E. Bercaw, J.
Am. Chem. Soc. 2002, 124, 1378.
[3] a) W. J. Tenn III, K. J. H. Young, G. Bhalla, J. Oxgaard, W. A.
Goddard III, R. A. Periana, J. Am. Chem. Soc. 2005, 127, 14172;
b) Y. Feng, M. Lail, K. A. Barakat, T. R. Cundari, T. B. Gunnoe,
J. L. Petersen, J. Am. Chem. Soc. 2005, 127, 14174.
[4] a) Y. Feng, M. Lail, N. A. Foley, T. B. Gunnoe, K. A. Barakat,
T. R. Cundari, J. L. Petersen, J. Am. Chem. Soc. 2006, 128, 7982;
b) W. J. Tenn III, K. J. H. Young, J. Oxgaard, R. J. Nielsen, W. A.
Goddard III, R. A. Periana, Organometallics 2006, 25, 5173.
[5] a) A. Sacco, G. Vasapollo, C. F. Nobile, A. Piergiovanni, M. A.
Pellinghelli, M. Lanfranchi, J. Organomet. Chem. 1988, 356, 397;
b) J. Zhang, G. Leitus, Y. Ben-David, D. Milstein, J. Am. Chem.
Soc. 2005, 127, 10840; c) J. Zhang, G. Leitus, Y. Ben-David, D.
Milstein, Angew. Chem. 2006, 118, 1131; Angew. Chem. Int. Ed.
2006, 45, 1113.
In conclusion, new RhI hydroxide, phenoxide, and tri-
fluoroethoxide complexes are shown to be capable of arene
C H bond activation. This reaction may proceed through a
À
À
mechanism similar to the two previous examples of C H
bond activation by high-valent transition-metal alkoxide/
hydroxide complexes,[3] but with the important distinction
that the low-valent RhI complexes reported herein do not
À
require loss of an ancillary ligand to facilitate the C H bond
[6] E. Ben-Ari, G. Leitus, L. J. W. Shimon, D. Milstein, J. Am. Chem.
Soc. 2006, 128, 15390.
[7] See the Supporting Information for details.
activation. It may be possible to expand these RhIOR
complexes to create a new general class of square-planar d8
À
metal complexes, which show high activity for this unusual C
[8] a) J. L. Garnett, R. J. Hodges, J. Am. Chem. Soc. 1967, 89, 4546;
b) J. L. Garnett, M. A. Long, K. B. Peterson, Aust. J. Chem. 1974,
27, 1823; c) M. R. Blake, J. L. Garnett, I. K. Gregor, W. Hannan,
K. Hoa, M. A. Long, J. Chem. Soc. Chem. Commun. 1975, 930;
d) A. E. Shilov, A. A. Shteinman, Coord. Chem. Rev. 1977, 24, 97;
e) S. R. Klei, J. T. Golden, D. T. Tilley, R. G. Bergman, J. Am.
Chem. Soc. 2002, 124, 2092; f) M. H. G. Prechtl, M. Holscher, Y.
Ben-David, N. Theyssen, R. Loschen, D. Milstein, W. Leitner,
Angew. Chem. 2007, 119, 2319; Angew. Chem. Int. Ed. 2007, 46,
2269; g) C. W. Leung, W. Zheng, D. Wang, S. M. Ng, C. H. Yeung,
Z. Zhou, Z. Lin, C. P. Lau, Organometallics 2007, 26, 1924.
[9] L. M. Stock, H. C. Brown, J. Am. Chem. Soc. 1959, 81, 3323.
H bond activation in the presence of water or alcohol. Future
work will focus on determining the mechanism of these
activation reactions and exploring the range of d8 square-
planar metal/ligand complexes that can promote this exciting
reactivity.
Received: January 19, 2007
Published online: May 10, 2007
À
Keywords: C H activation · H/D exchange · hydroxide ligands ·
.
NMR spectroscopy · rhodium
4738
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4736 –4738