Slow reductive elimination from arylpalladium parent amido complexes

Article abstract of DOI:10.1021/ja1023404

We report reductive eliminations of primary arylamines from a series of bisphosphine-ligated arylpalladium(II) parent amido complexes that counter several established trends. In contrast to arylamido and alkylamido complexes of the aromatic bisphosphines

Full text of DOI:10.1021/ja1023404

Published on Web 08/10/2010
Slow Reductive Elimination from Arylpalladium Parent Amido Complexes
Jessica L. Klinkenberg and John F. Hartwig*
Department of Chemistry, UniVersity of Illinois, 600 South Mathews AVenue, Urbana, Illinois 61801
Received March 19, 2010; E-mail: jhartwig@uiuc.edu
1. These complexes were generated by combining arylpalladium
bromide precursors, (CyPF-t-Bu)Pd(Ar)(Br) (Ar ) p-tolyl, p-anisyl,
o-tolyl, o-anisyl), 1 atm of ammonia, and 1.4 equiv of NaO-t-Bu
base. Conversion of 1a-1d to the amido complexes occurred in
Abstract: We report reductive eliminations of primary arylamines
from a series of bisphosphine-ligated arylpalladium(II) parent
amido complexes that counter several established trends. In
contrast to arylamido and alkylamido complexes of the aromatic
bisphosphines DPPF and BINAP, parent amido complexes of
these ligands do not form or undergo reductive elimination of
monoarylamines. However, arylpalladium parent amido com-
plexes ligated by the alkylbisphosphine CyPF-t-Bu do form in
less than 1 h at 25 °C. These complexes were found to be stable
enough at -30 °C to be characterized in solution by low temperature
1H and 31P NMR spectroscopy and infrared spectroscopy, including
data on 15N-labeled analogs.
good yield and undergo reductive elimination. Despite the basicity
of the parent amido ligand and the typically faster reductive
elimination from complexes containing more basic amido ligands,
the CyPF-t-Bu-ligated arylpalladium parent amido complexes
undergo reductive elimination much more slowly than the analo-
gous complexes containing arylamido or alkylamido ligands.
Moreover, the parent amido complexes form more rapidly and
are more stable thermodynamically in a series of exchange
processes than the arylamido complexes. Computational studies
support the overriding influence of steric effects on the stability
and reactivity of the parent amido complex. The slow rate of
reductive elimination causes the arylpalladium amido complex to
be the resting state of the coupling of aryl halides with ammonia
catalyzed by CyPF-t-Bu-ligated palladium, and this resting state
contrasts the Pd(0) or arylpalladium(II) resting states of reactions
of aryl halides with amines catalyzed by most palladium complexes.
Scheme 1. In Situ Generation of Parent Amido Complexes 2a-d
Arylpalladium arylamido and alkylamido complexes containing
CyPF-t-Bu as ligand were also generated to compare their reactivity
to that of complexes 2a-2d. Arylamido complexes 3a and 3b
formed in ∼70 and 60% yield, respectively, as shown in Scheme
2 and were characterized by NMR spectroscopy at -30 °C.
Although we were able to characterize arylpalladium parent amido
complexes containing ortho substituents on the palladium bound
aryl group, arylpalladium anilido complexes containing ortho
substituents did not form at room temperature. The slow formation
of such ortho-substituted arylpalladium arylamido complexes is
consistent with selective formation of primary vs secondary
arylamines from the coupling of ammonia with ortho-substituted
aryl halides. Reaction of complex 1a with iso-butylamine and
NaO-t-Bu at -30 °C did not lead to observable quantities of an
alkylamido complex; instead, the N-alkyl aniline product (>90%)
and (CyPF-tBu)Pd(PPh3) (>98%), which would form by reductive
elimination from the arylpalladium iso-butylamido complex in the
presence of PPh3, were the only species observed.
Parent amido complexes of the late transition metals are rare,
and few classes of catalytic or stoichiometric reactions have been
discovered that occur through such species. The existing stoichio-
metric reactions demonstrate that this class of compound is highly
basic at nitrogen,1-3 but studies of reaction chemistry have been
limited to proton exchanges with weak acids,1-6 insertions of CO
into the N-H bond,7 and reductive eliminations of ammonia.8 The
recently discovered palladium-catalyzed coupling of aryl halides
with ammonia is likely to occur through parent amido complexes,9-13
but only one potential amido intermediate in this process has been
generated, and it was investigated briefly.5a Thus, no study reveals
the relative thermodynamic stability and reactivity of parent amido
complexes versus the analogous alkyl- and arylamido complexes
containing the same ancillary ligands.
Scheme 2. Generation of Arylpalladium Arylamido Complexes
3a-b
We report results from studies of a series of arylpalladium amido
complexes containing bisphosphine ligands that contradict several
established trends. The parent amido complexes we generate are
more stable thermodynamically than amides derived from more
acidic amines, undergo reductive elimination when ligated by only
a select set of bisphosphines, and reductively eliminate more slowly
than complexes containing less basic arylamido and more basic
alkylamido ligands. They are sufficiently stable that the arylpalla-
dium amido complex is even the resting state of the catalyst bound
by a hindered alkylbisphosphine ligand.
Reductive eliminations of arylamines from arylamido and
alkylamido complexes ligated by DPPF have been reported
previously,14-16 and similar reactions of arylpalladium amido
complexes ligated by BINAP certainly occur during reactions
catalyzed by palladium complexes of this ligand.17 In contrast to
these systems containing alkyl- and arylamido ligands, DPPF- and
Our synthesis of arylpalladium parent amido complexes 2a-2d
containing the Josiphos ligand CyPF-t-Bu is summarized in Scheme
9
11830 J. AM. CHEM. SOC. 2010, 132, 11830–11833
10.1021/ja1023404  2010 American Chemical Society

C O M M U N I C A T I O N S
Scheme 3. Reductive Elimination of Arylamines from Parent Amido
Complexes 2a-d and Arylamido Complexes 3a-b
BINAP-ligated arylpalladium parent amido complexes did
not undergo reductive elimination of arylamines. Reaction of
(L)Pd(Ar)(Br) (L ) DPPF, BINAP, Ar ) C6H4-p-OMe, Ph) with
ammonia and NaO-t-Bu did not lead to (L)Pd(Ar)(NH2) or primary
arylamines; these reactions led to homocoupling of the aryl ligand
and unidentified Pd side products. The alternative route to a BINAP-
ligated parent amido complex involving deprotonation of
[(BINAP)Pd(Ph)(NH3)](OTf)18 led to a new species by 31P NMR
that formed biphenyl products, free PPh3, and less than 5% of
diphenylamine. Consistent with these results, the combination of a
Pd(0) precatalyst and DPPF or BINAP does not catalyze the
monoarylation of ammonia under the previously described condi-
tions for reactions catalyzed by CyPF-t-Bu-ligated palladium.5
a
These data with two common arylphosphines imply that arylpal-
ladium parent amido complexes undergo reductive elimination more
slowly than arylpalladium alkylamido and arylamido complexes,
are more open to competing side reactions, or both.
In contrast, both parent amido complexes and arylamido com-
plexes ligated by CyPF-t-Bu undergo reductive elimination, as
shown in Scheme 3. In the presence of 2 equiv of PPh3 to trap the
Pd(0) product, the complexes ligated by CyPF-t-Bu formed the
primary arylamines in 50-86% yield for the two-step sequence at
room temperature; the reductive elimination step formed the
arylamine in 60-90% yield and the Pd(0) species in 55-85% yield.
The aniline products formed in higher yields when they were
allowed to undergo reductive elimination at elevated temperatures,
the largest difference being observed for the reductive elimination
of p-toluidine from 2b.
Reactions of the arylpalladium anilides 3a and 3b occurred
rapidly at room temperature to form diarylamine products in
36-53% yield for the two-step sequence and 60-70% for the
reductive elimination step. This fast reductive elimination from the
arylamido complex at room temperature, along with the fast
formation of the N-alkyl arylamine from reaction of 1a with iso-
butylamine and NaO-t-Bu, shows qualitatively that reductive
elimination from the parent amido complex is slower than from
both more electron-poor arylamido complexes and more electron-
rich alkylamido complexes.
Rate constants for reductive eliminations from the parent amido
and arylamido complexes in the presence of 2 equiv of PPh3
determined by 31P NMR spectroscopy at room temperature are
shown in Scheme 3.19 Reactions conducted with 2, 4, and 8 equiv
of added PPh3 occurred with indistinguishable rate constants,
indicating that the reductive elimination is zero-order in added
phosphine and occurs directly from the arylpalladium amide.
The data in Scheme 3 reveal large differences in the rates for
reductive elimination from the various arylpalladium parent amido
complexes. The rate constant for reductive elimination from the
para-anisylpalladium complex 2a was an order of magnitude
smaller than that from the less electron-rich para-tolyl complex
2b. Complexes 2c and 2d containing an ortho substituent on the
aryl group underwent reductive elimination nearly 50 times faster
than those containing para substituents.
a GC yield of monoarylamine or diarylamine at 20 °C for the reductive
elimination step. b At 20 °C. c Determined by monitoring the decay of the
amido or arylamido complex by 31P NMR spectroscopy. d THF/dioxane
mixture of 3:2. e Average of 2-4 runs. A standard deviation is provided
for the cases that are run more than twice and have the largest errors; the
rate constants for cases run twice are within 5% of each other. f Amido
complex generated with KO-t-Bu as the base. g Amido complex generated
with LiO-t-Bu as the base.
complexes are known to be highly basic,1-3 the smaller size of
the parent amido ligand is, most likely, responsible for the large
differences in the rates of reductive elimination.
To understand the thermodynamic properties of the different
amido complexes and origins of the relative rates for reductive
elimination, we assessed the different thermodynamic stabilities of
the parent amido and arylamido complexes, relative to the free
amines, by both computational and experimental methods, and we
computed the barriers to reductive elimination from a series of
arylpalladium amido complexes. To distinguish between steric and
electronic effects on the stability and reactivity of the complexes
containing the different amido groups, we assessed by DFT
calculations the thermodynamic properties and barriers for reductive
eliminations involving complexes in which steric effects should
have less of an influence on the relative thermodynamic stability
of the complexes and rates for reductive elimination than they do
in the experimental system.
A quantitative assessment of the relative rates for reductive
elimination from complexes containing different types of amido
groups is also contained in Scheme 3. The para-anisylamido
complex 3a reacted 100 times faster than the corresponding parent
amido complex 2a, and the para-tolylamido complex 3b reacted
150 times faster than the corresponding parent amido complex 2b,
the half-life for reaction of 3b being less than a minute. Because
we have shown previously that amido complexes containing more
basic nitrogen atoms undergo faster reductive eliminations than
those with less basic nitrogen atoms,15 and terminal amido
The stability of the amido complexes, transition states for C-N
reductive elimination, and overall free energies for formation of
the amine product in dioxane solvent were calculated for a series
of cis-bis(trimethylphosphine)-ligated phenylpalladium amido com-
plexes, and the results of this analysis are shown in Figure 1. As
expected from the greater acidity of aniline,11 the equilibrium
between the combination of the arylamido complex and ammonia
and the combination of the parent amido complex and aniline
favored the arylamido complex by a free energy of 1.9 kcal/mol.
9
J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010 11831

C O M M U N I C A T I O N S
Figure 1. Relative ground state and transition state free energies in kcal/
mol for PMe3-ligated phenylpalladium parent amido, arylamido, and
methylamido complexes undergoing C-N reductive elimination.
The free energy barrier for reductive elimination of diphenylamine
from the anilide complex was calculated to be higher than that for
reductive elimination of aniline from the parent amido complex
by just 0.3 kcal/mol. The reductive elimination of aniline was more
downhill than the reductive elimination of diphenylamine by 3.0
kcal/mol in free energy. The N-methylamido complex was calcu-
lated to be significantly less stable than either the parent amido or
arylamido complex in the exchange equilibria, but the barrier for
reductive elimination from the methylamido complex was calculated
to be more than 4 kcal/mol lower than that for reductive elimination
from either the parent amido or arylamido complexes.
Scheme 4. Mechanism of Pd-Catalyzed Amination with Ammonia
and Observed Arylpalladium Parent Amido Resting State
The calculated thermodynamic values can then be compared to
experimental results with the arylpalladium amido complexes ligated
by more sterically encumbered CyPF-t-Bu. In contrast to the typical
trend of the base of a more acidic molecule being the preferred
ligand among an analogous series of acid-base pairs20 (as
underscored by the computed equilibria for the PMe3 complexes),
the palladium center ligated by CyPF-t-Bu favored binding of the
parent amido group. Although competitive reductive elimination
from the arylpalladium arylamido complexes prevented our obtain-
ing quantitative Keq values for the proton transfer equilibria, reaction
of complex 1b with ammonia and p-toluidine in the presence of
1.4 equiv of NaO-t-Bu generated a mixture containing a higher
population of parent amide 2b than of the arylamide 3b (eq 1).
Moreover, treatment of arylamide 3a generated in situ with 5 equiv
of ammonia began to form 2a (1:2.5 ratio of 2a to 3a after 20 min
at -40 °C), but treatment of 2a with 5 equiv of p-anisidine at -40
°C gave little arylamide 3a (ratio of 2a to 3a was >20:1 after 20
min). When the latter sample was warmed to room temperature
and allowed to equilibrate for an additional 20 min, parent amide
2a was favored over arylamide 3a by a ratio of 3:1 (eq 2). Thus,
the faster generation of the parent amide 2a and greater thermo-
dynamic stability of 2a lead to selective formation of the primary
arylamine, not faster reductive elimination from the parent amido
complexes due to the strong basicity of the parent amido group.1,2,21
At the same time, the strong basicity of the primary alkylamido
complexes does appear to affect the rates of reductive elimination.
The low barrier computed for reductive elimination from the
methylamido complex agrees with the fast reductive elimination
observed from the CyPF-tBu-ligated isobutylamido complex gener-
ated in situ.
dium amide by ammonia and NaO-t-Bu leads to the arylpalladium
parent amido complex. Reductive elimination from this species
forms the monoarylamine. Monitoring of the reaction of 4-bromo-
toluene with ammonia in dioxane by 31P NMR spectroscopy at 80
°C with 11 mol % of the catalyst showed that the parent amido
complex was the major species present in the reaction solution.
This amide resting state contrasts with the Pd(0) resting state in
reactions of amines with aryl halides catalyzed by complexes of
DPPF and BINAP.17,22 Consistent with the parent amido complex
as the resting state, the rate constant for the catalytic coupling at
50 °C with 5 mol % of the preformed arylpalladium halide complex
1b was similar (8.0 × 10-4 vs 7.6 × 10-4 s-1) to that for the
stoichiometric reaction of the arylpalladium parent amido complex
at the same temperature, after correcting for the catalyst concentra-
tion (Scheme 4).
In summary, the chemistry of the parent amido complexes
[Pd(CyPF-t-Bu)(Ar)(NH2)], as well as analogs containing aromatic
bisphosphine ligands, is distinct from that of the alkyl- and
arylamido analogs. The parent amido complexes ligated by CyPF-
t-Bu are more stable thermodynamically and undergo reductive
elimination more slowly than the alkylamido and arylamido analogs.
This difference from the usual trends in stability and reactivity that
result from electronic effects is likely due to the hindered ligand
environment of CyPF-t-Bu. This hindered environment disfavors
formation of complexes of the larger amides and reduces the
energies for reductive elimination from complexes of the bulkier
aryl- and alkyl-substituted amides. Finally, parent amido complexes
containing common aromatic bisphosphine ligands do not form or
undergo reductive elimination of arylamines in high yields.
Finally, we studied the identity of the palladium complex in the
coupling of aryl halides with ammonia catalyzed by the combination
of Pd[P(o-tol)3]2 and CyPF-t-Bu. The mechanism for this reaction
is shown below in Scheme 4. Like many common C-N cross-
coupling reactions, the process occurs by initial oxidative addition
of an aryl halide to the Pd(0) precursor (CyPF-t-Bu)Pd(P-o-tol3).
Conversion of the arylpalladium halide complex to the arylpalla-
Acknowledgment. We thank the NIH (R37-GM-55382) for
support, Johnson-Matthey for Pd, and Solvias for CyPF-t-Bu.
Supporting Information Available: All experimental procedures
and spectroscopic data of new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
9
11832 J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010

C O M M U N I C A T I O N S
(16) Yamashita, M.; Vicario, J. V. C.; Hartwig, J. F. J. Am. Chem. Soc. 2003,
125, 16347–16360.
References
(17) Alcazar-Roman, L. M.; Hartwig, J. F.; Rheingold, A. L.; Liable-Sands,
L. M.; Guzei, I. A. J. Am. Chem. Soc. 2000, 122, 4618–4630.
(1) Fulton, J. R.; Bouwkamp, M. W.; Bergman, R. G. J. Am. Chem. Soc. 2000,
122, 8799–8800.
(18) See the Supporting Information for the preparation of this species.
(2) Fulton, J. R.; Sklenak, S.; Bouwkamp, M. W.; Bergman, R. G. J. Am. Chem.
Soc. 2002, 124, 4722–4737.
(19) A reviewer suggested that the cation of the base used to generate the amido
complexes might bind to the parent amide and retard reductive elimination
of primary arylamines. Thus, we measured the rates for reductive
elimination of 2b formed from Li and KO-t-Bu. These rates were
indistinguishable, and the similar rates imply that the cation is not bound
to the amido ligand in the starting material or transition state.
(3) Jayaprakash, K. N.; Conner, D.; Gunnoe, T. B. Organometallics 2001, 20,
5254–5256.
(4) Fox, D. J.; Bergman, R. G. Organometallics 2004, 23, 1656–1670.
(5) Conner, D.; Jayaprakash, K. N.; Cundari, T. R.; Gunnoe, T. B. Organo-
metallics 2004, 23, 2724–2733.
(6) Conner, D.; Jayaprakash, K. N.; Wells, M. B.; Manzer, S.; Gunnoe, T. B.;
Boyle, P. D. Inorg. Chem. 2003, 42, 4759–4772.
(20) Holland, P. L.; Andersen, R. A.; Bergman, R. G. Comm. Inorg. Chem.
1999, 21, 115–129.
(7) Fox, D. J.; Bergman, R. G. J. Am. Chem. Soc. 2003, 125, 8984–8985.
(8) Kanzelberger, M.; Zhang, X.; Emge, T. J.; Goldman, A. S.; Zhao, J.;
Incarvito, C.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125, 13644–13645.
(9) Shen, Q.; Hartwig, J. F. J. Am. Chem. Soc. 2006, 128, 10028–10029.
(10) Vo, G. D.; Hartwig, J. F. J. Am. Chem. Soc. 2009, 131, 11049–11061.
(11) Surry, D. S.; Buchwald, S. L. J. Am. Chem. Soc. 2007, 129, 10354–10355.
(12) Schulz, T.; Torborg, C.; Enthaler, S.; Schaffner, B.; Dumrath, A.;
Spannenberg, A.; Neumann, H.; Borner, A.; Beller, M. Chem.sEur. J. 2009,
15, 4528–4533.
(21) Preliminary data suggest that the establishment of Curtin-Hammett
conditions for the equilibrium between the two amido complexes depends
on temperature. At higher temperature, the rates of reductive elimination
increase more than the rates for intermolecular exchanges. More detailed
studies on the precise origins of selectivity for the coupling of ammonia
will be the subject of future work.
(22) Shekhar, S.; Ryberg, P.; Hartwig, J. F.; Mathew, J. S.; Blackmond,
D. G.; Strieter, E. R.; Buchwald, S. L. J. Am. Chem. Soc. 2006, 128,
3584–3591.
(13) Lundgren, R. J.; Sappong-Kumankumah, A.; Stradiotto, M. Chem.sEur.
J. 2010, 16, 1983–1991.
(14) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 7217–7218.
(15) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1997, 119, 8232–8245.
JA1023404
9
J. AM. CHEM. SOC. VOL. 132, NO. 34, 2010 11833