Unusually stable palladium(IV) complexes: Detailed mechanistic investigation of C-O bond-forming reductive elimination

Article abstract of DOI:10.1021/ja0541940

This communication describes the synthesis of a family of unusually stable palladium(IV) complexes containing two chelating 2-phenylpyridine ligands and two benzoates. These complexes undergo clean C-O bond-forming reductive elimination upon heating, and the mechanism of this catalytically relevant process has been studied in detail. Solvent effects, crossover experiments, Eyring plots (which show ΔS_≠ of -1.4 ± 1.9 and 4.2 ± 1.4 in CDCl₃ and DMSO, respectively), and Hammett analysis (which shows ρ = -1.36 ± 0.04 upon substitution of the para-benzoate substituent) all suggest that reductive elimination does not proceed via initial dissociation of a benzoate ligand. Instead, an unusual mechanism involving pre-equilibrium dissociation of the N-arm of the phenylpyridine ligand is proposed. Copyright

Full text of DOI:10.1021/ja0541940

Published on Web 08/23/2005
Unusually Stable Palladium(IV) Complexes: Detailed Mechanistic
Investigation of C-O Bond-Forming Reductive Elimination
Allison R. Dick, Jeff W. Kampf, and Melanie S. Sanford*
Department of Chemistry, UniVersity of Michigan, 930 North UniVersity AVenue, Ann Arbor, Michigan 48109
Received June 24, 2005; E-mail: mssanfor@umich.edu
Carbon-oxygen bond-forming reductive elimination is a fun-
damental organometallic transformation involved in a wide variety
of transition metal catalyzed processes.1 Of particular current
relevance, C-O bond-forming reductive elimination from transient
PdIV aryl/acetate intermediates has been implicated as the product
release step in PdII-catalyzed arene oxygenation reactions.2 Recent
elegant studies have probed the mechanisms of related C-O bond-
4
forming processes at NiIII,1d PdII,3 and PtIV centers, uncovering
distinctly different pathways in each system. In contrast, a detailed
mechanistic picture of C-O bond formation from PdIV has remained
elusive.5 Attempts to investigate this process have been hampered
by the low stability of PdIV complexes (particularly those containing
Figure 1. ORTEP diagram of 2k.
multiple oxygen donor ligands) as well as by the propensity of PdIV
Remarkably, complex 2a is stable in the solid state for at least
to undergo undesired side reactions, such as competing C-C bond-
a week and shows little decomposition after an hour in CDCl3
forming reductive elimination and/or intermolecular alkyl ligand
solution at 25 °C.7 (In contrast, analogous (C2N2O2)PdIV complexes
exchange.5-7 We report herein the rational design and synthesis
that do not contain cyclometalated ligands have been implicated
of a series of unusual PdIV complexes (2a-2o) that are remarkably
as reactive intermediates but not detected even at -70 °C.)5a
stable at ambient temperature but undergo clean C-O bond-
However, we were pleased to find that 2a does undergo smooth
forming reductiVe elimination upon thermolysis. These complexes
C-O bond-forming reductive elimination to afford 4a when heated
have enabled the first detailed mechanistic investigation of C-O
for 1 h at 60 °C (eq 2). This reaction proceeds with clean first-
bond-forming reductive elimination from PdIV.
order kinetics when conducted in the presence of 5% d5-pyridine
In designing PdIV model complexes for these studies, we sought
(added to trap putative three-coordinate PdII intermediate 3a),10
to incorporate features that would both stabilize the +4 oxidation
facilitating detailed mechanistic studies of this process.
state and promote C-O bond formation over competing side
reactions. In this regard, our first synthetic target was complex 2a
(eq 1), containing two rigid cyclometalated pyridine ligands (to
stabilize PdIV)5-8 with aryl rather than alkyl groups (to prevent
ligand exchange between metal centers).5-8 These aryl ligands were
also placed in independent rigid chelating frameworks, which we
reasoned should reduce undesired ligand exchange and C-C bond-
forming processes. Finally, benzoate-based O-donors were incor-
porated to allow facile manipulation of electronic parameters
(through variation of the para-substituent) as well as to model Pd-
catalyzed arene oxygenation reactions.2
As summarized in Scheme 1, we considered three distinct
mechanistic pathways for reductive elimination from 2as(i) pre-
equilibrium dissociation of a benzoate ligand followed by either
external or intramolecular nucleophilic attack (A), (ii) direct
reductive elimination from the six-coordinate starting material (B),
or (iii) dissociation of a pyridyl arm of one cyclometalated ligand
followed by internal coupling (C). We initially felt that mechanism
A was most likely, by analogy to related C-O bond-forming
reactions at PtIV (which proceed by pre-equilibrium dissociation
of RO-)4 and to C-C coupling at PdIV (which involves pre-
equilibrium loss of I-).11 As a result, we first examined the effect
of solvent on the rate of reductive elimination from 2o (a more
-
Our studies began with complex 1, which was readily prepared
according to a literature procedure.9 Gratifyingly, we found that 1
undergoes clean oxidation with PhI(O2CPh)2 to afford the isolable
soluble analogue of 2a containing two C9H19CO2 ligands),
expecting a significant acceleration in more polar solvents.4,11
However, surprisingly, the reaction proceeded at essentially identical
rates in polar acetone (ꢀ ) 21, krel ) 1.0 ( 0.1) and nonpolar
benzene (ꢀ ) 2.3, krel ) 1.0 ( 0.1) at 55 °C (krel ) kobs/kacetone).
Furthermore, the rate showed no discernible correlation with solvent
dielectric constant: DMSO (ꢀ ) 47, krel ) 2.0 ( 0.3) < CHCl3 (ꢀ
) 4.8, krel ) 2.3 ( 0.2) < MeCN (ꢀ ) 38, krel ) 2.4 ( 0.1) <
nitrobenzene (ꢀ ) 36, krel ) 3.1 ( 0.3).12 In contrast, the rates of
1
bisbenzoate PdIV complex 2a in 77% yield (eq 1). The H NMR
spectrum of 2a shows 22 inequivalent aromatic resonances,
consistent with an asymmetric cis-geometry with free rotation about
the C-Ph bond of the benzoate ligands. The solid-state structure
of the para-NO2 analogue (2k) was further confirmed by X-ray
crystallography (Figure 1).
9
12790
J. AM. CHEM. SOC. 2005, 127, 12790-12791
10.1021/ja0541940 CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Scheme 1. Possible Mechanisms of C-O Reductive Elimination
6 took 2.5 h to reach completion. On the basis of this large
(approximately 40-fold) difference, we currently favor mechanism
C for these reactions, and further studies to confirm this hypothesis
are underway.
In summary, we have demonstrated the design and synthesis of
a series of remarkably stable PdIV complexes and have presented
the first detailed study of C-O bond-forming reductive elimination
from this oxidation state. These experiments indicate that C-O
coupling at PdIV proceeds by a significantly different mechanism
than other reductive eliminations from PdIV or PtIV centers. Current
work in our laboratory aims to exploit these mechanistic insights
for the development of new PdIV-catalyzed reactions.
ionic reductive elimination from both PtIV and PdIV typically show
strong dependence on solvent polarity.4,11
A series of additional experiments provided further evidence
against benzoate dissociation mechanism A. First, Eyring analysis
of reductive elimination from 2c afforded ∆Sq values of +4.2 (
1.4 and -1.4 ( 1.9 eu in d6-DMSO and CDCl3, respectively.12 In
contrast, ionic reductive elimination reactions typically show highly
negative values of ∆Sq as a result of solvent ordering about the
charged transition state. (For example, ∆Sq values ranging from
-13 to -49 eu are typical of ionic C-C and C-Se reductive
elimination from PdIV.)6,11 Additionally, the rate of reductive
elimination was examined in a series of complexes containing para-
substituted benzoate ligands (2b-2k). Electron donor substituents
led to moderate rate accelerations [with a Hammett F value of -1.36
( 0.04 (R2 ) 0.98)], indicating that the benzoate acts as a
nucleophilic partner in these transformations.12 A comparable F
value of -1.5 has been reported for C-S coupling at PdII, which
is believed to proceed by a mechanism similar to B.10 In contrast,
C-O bond-forming reductive elimination from PtIV (which proceeds
via mechanism A) shows a F of +1.44, indicative of stabilization
of the dissociated RO- moiety by electron-withdrawing groups.4
Finally, thermolysis of mixtures of 2l and 2g (two differentially
substituted PdIV complexes that reductively eliminate at comparable
rates) yielded oxygenated organic products without any observable
crossover in CHCl3 or DMSO.13 Furthermore, thermolysis of 2b
in the presence of 5 equiv of NBu4OAc resulted in e5%
incorporation of acetate into the organic reductive elimination
product in CHCl3 or DMSO. These results provide further evidence
against mechanism A as the major reaction pathway since extensive
exchange between ion pairs and/or free ions would be expected if
benzoate dissociation preceded reductive elimination in these
systems.
In sum, these studies led us to the surprising conclusion that
C-O bond-forming reductive elimination from complex 2a pro-
ceeds predominantly by either direct reductive elimination from
the octahedral starting material (a rare process in both PtIV and
PdIV chemistry)14,15 or by dissociation of an arm of one of the
chelating phenylpyridine ligands (mechanisms B and C, respec-
tively). These mechanisms are kinetically indistinguishable and
cannot be definitively differentiated based on any of the experiments
detailed above. However, we reasoned that preliminary evidence
to distinguish B and C might be obtained by comparing the rate of
reductive elimination from bisphenylpyridine complex 2o to that
from bisbenzo[h]quinoline complex 6 (eq 3). In the case of
mechanism B, comparable rates of reductive elimination are
expected for 2o and 6, due to the similar steric and electronic
parameters of the ligands. However, in the case of mechanism C,
the added rigidity of the fused ring system is expected to
dramatically decrease the rate of nitrogen dissociation and, therefore,
the overall rate of reductive elimination from 6 relative to 2o.16
When these complexes were heated at 75 °C in CD3CN, reductive
elimination from 2o was complete in 4 min, while the reaction of
Acknowledgment. We thank the Camille and Henry Dreyfus
Foundation, the Arnold and Mabel Beckman Foundation, and Eli
Lilly (graduate fellowship to A.R.D.) for support. We also thank
Dr. E. Alvarado and Dr. C. Kojiro for assistance with NMR kinetics.
Supporting Information Available: Experimental details, crystal-
lographic data for 2k, spectroscopic and analytical data for new
compounds, and detailed discussion of kinetics (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002, 219, 131. (b)
Hartwig, J. F. Acc. Chem. Res. 1998, 31, 852. (c) Stahl, S. S.; Labinger,
J. A.; Bercaw, J. E. Angew. Chem., Int. Ed. 1998, 37, 2181. (d) Han, R.;
Hillhouse, G. L. J. Am. Chem. Soc. 1997, 119, 8135.
(2) For recent examples, see: (a) Desai, L. V.; Hull, K. L.; Sanford, M. S. J.
Am. Chem. Soc. 2004, 126, 9542. (b) Dick, A. R., Hull, K. L.; Sanford,
M. S. J. Am. Chem. Soc. 2004, 126, 2300. (c) Yoneyama, T.; Crabtree,
R. H. J. Mol. Catal. A 1996, 108, 35.
(3) (a) Mann, G.; Shelby, Q.; Roy, A. H.; Hartwig, J. F. Organometallics
2003, 22, 2775. (b) Widenhoefer, R. A.; Buchwald, S. L. J. Am. Chem.
Soc. 1998, 120, 6504. (c) Widenhoefer, R. A.; Zhong, H. A.; Buchwald,
S. L. J. Am. Chem. Soc. 1997, 119, 6787.
(4) Williams, B. S.; Goldberg, K. I. J. Am. Chem. Soc. 2001, 123, 2576.
(5) (a) Canty, A. J.; Denney, M. C.; Skelton, B. W.; White, A. H.
Organometallics 2004, 23, 1122. (b) Canty, A. J.; Denney, M. C.; van
Koten, G.; Skelton, B. W.; White, A. H. Organometallics 2004, 23, 5432.
(c) Canty, A. J.; Jin, H. J. Organomet. Chem. 1998, 565, 135. (d) Valk,
J.-M.; Boersma, J.; van Koten, G. Organometallics 1996, 15, 4366.
(6) For preliminary studies of C-Se reductive elimination, see: Canty, A.
J.; Jin, H.; Skelton, B. W.; White, A. H. Inorg. Chem. 1998, 37, 3975.
(7) A related stable C2O2 PdIV complex has been reported; however,
thermolysis afforded a complex mixture, preventing detailed mechanistic
studies of reductive elimination. Yamamoto, Y.; Kuwabara, S.; Matsuo,
S.; Ohno, T.; Nishiyama, H.; Itoh, K. Organometallics 2004, 23, 3898.
(8) Alsters, P. L.; Engel, P. F.; Hogerheide, M. P.; Copijn, M.; Spek, A. L.;
van Koten, G. Organometallics 1993, 12, 1831.
(9) Jolliet, P.; Gianini, M.; von Zelewsky, A.; Bernardinelli, G.; Stoeckli-
Evans, H. Inorg. Chem. 1996, 35, 4883.
(10) Ligands are frequently added to trap unsaturated metal fragments after
reductive elimination. See: Mann, G,; Baranano, D.; Hartwig, J. F.;
Rheingold, A. L.; Guzei, I. A. J. Am. Chem. Soc. 1998, 120, 9205.
(11) Byers, P. K.; Canty, A. J.; Crespo, M.; Puddephatt, R. J.; Scott, J. D.
Organometallics 1988, 7, 1363.
(12) Studies were conducted in the presence of 5% C5D5N to obtain clean
kinetics; qualitatively, the rates with and without C5D5N were similar.
(13) The lack of crossover in the organic products was determined by 1H NMR
1
and GCMS. Overlapping H NMR signals prevented definitive determi-
nation of whether there was crossover in the inorganic products.
(14) For rare direct reductive elimination from PtIV, see: (a) Crumpton-Bregel,
D. M.; Goldberg, K. I. J. Am. Chem. Soc. 2003, 125, 9442. (b) Edelbach,
B. L.; Lachicotte, R. J.; Jones, W. D. J. Am. Chem. Soc. 1998, 120, 2843.
(15) C-C reductive elimination from PdIV typically proceeds by mechanism
A, with minimal (<5%) contribution from B or C (ref 11).
(16) Related experiments have been used to differentiate mechanisms involving
dissociation of chelating phosphines (see refs 10 and 14a).
JA0541940
9
J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005 12791