Catalysis with platinum-group alkylamido complexes. The active palladium amide in catalytic aryl halide aminations as deduced from kinetic data and independent generation

Article abstract of DOI:10.1021/om960188o

Mechanistic studies of the palladium-catalyzed coupling between aryl bromides and tin amides were conducted as a means to evaluate the pathway of this reaction as well as the general potential of low valent amido complexes to be reactive intermediates in catalysis. The specific systems involved reactions between Bu₃SnNMe₂ and aryl halides catalyzed by {Pd[P(o-Tol)₃]₂} (1), {Pd[P(o-Tol)₃](P-MeC₆H₄)(Br)}₂ (2a), and {Pd[P(o-Tol)₃](NHMe₂)(p-MeCeH₄)(Br)} (3a). A combination of kinetic studies and independent synthesis of reaction intermediates indicated that the three-coordinate platinum-group amido complex {Pd[P(o-Tol)₃](Ar)(NMe₂)} was an intermediate in these reactions. Thus, these aryl halide animations are rare examples of catalysis with a platinum-group amido complex. Kinetic data were obtained by ¹H NMR spectroscopy, and the rate behavior was determined to be zero order in added phosphine, zero order in aryl halide, and first order in tin amide under conditions of equal or greater concentrations of aryl bromide compared to tin amido. Reactions catalyzed by 3a were first order in the palladium complex. Reaction rates were inhibited by added tin bromide, but not by the arylamine product. The inhibition by tin bromide showed that reversible transmetalation between an aryl halide complex and the tin reagent was occurring. Subsequent to reversible transmetalation, a rate-determining reductive elimination of arylamine occurred. Under conditions with a 10-fold excess of tin amide and high phosphine concentrations, the rate-determining-step became oxidative addition of aryl bromide, and reactions became first order, rather than zero order, in aryl bromide. The amido intermediate deduced from kinetic studies appeared to be generated by reacting {Pd[P(o-Tol)₃](P-BuC₆H₄)-(Br}₂ (2b) with lithium arylamides or by deprotonating {Pd[P(o-Tol)₃](NHEt₂)(P-BuC₆H ₄)-(Br)} (3b) with MN(SiMe₃)₂ (M = K, Li). Both reactions gave yields of arylamine that were comparable to those of catalytic reactions. Competition and relative rate studies revealed an equilibrium between aryl halide complexes 2a-c and a tin amide adduct of it. In competition studies involving an in situ selectivity for reaction of Bu₃SnNMe₂ or Bu₃SnNEt₂ with p-t-BuC₆H₄Br catalyzed by 1, the ratio of N,N-dimethylaniline to N,N-diethylaniline was 2.9. However, kinetic measurements of individual reactions showed that Bu₃SnNMe₂ reacted only 1.4 times faster than Bu₃SnNEt₂, consistent with a reversible equilibrium involving tin amide binding to the catalyst, similar to that resulting from substrate binding pre?quilibria in enzyme systems.

Full text of DOI:10.1021/om960188o

2794
Organometallics 1996, 15, 2794-2805
Ca ta lysis w ith P la tin u m -Gr ou p Alk yla m id o Com p lexes.
Th e Active P a lla d iu m Am id e in Ca ta lytic Ar yl Ha lid e
Am in a tion s As Ded u ced fr om Kin etic Da ta a n d
In d ep en d en t Gen er a tion
J anis Louie, Fre´de´ric Paul, and J ohn F. Hartwig*
Department of Chemistry, Yale University, P.O. Box 208107,
New Haven, Connecticut 06520-8107
Received March 12, 1996X
Mechanistic studies of the palladium-catalyzed coupling between aryl bromides and tin
amides were conducted as a means to evaluate the pathway of this reaction as well as the
general potential of low valent amido complexes to be reactive intermediates in catalysis.
The specific systems involved reactions between Bu3SnNMe2 and aryl halides catalyzed by
{Pd[P(o-Tol)3]2} (1), {Pd[P(o-Tol)3](p-MeC6H4)(Br)}2 (2a ), and {Pd[P(o-Tol)3](NHMe2)(p-
MeC6H4)(Br)} (3a ). A combination of kinetic studies and independent synthesis of reaction
intermediates indicated that the three-coordinate platinum-group amido complex {Pd[P(o-
Tol)3](Ar)(NMe2)} was an intermediate in these reactions. Thus, these aryl halide aminations
are rare examples of catalysis with a platinum-group amido complex. Kinetic data were
1
obtained by H NMR spectroscopy, and the rate behavior was determined to be zero order
in added phosphine, zero order in aryl halide, and first order in tin amide under conditions
of equal or greater concentrations of aryl bromide compared to tin amide. Reactions catalyzed
by 3a were first order in the palladium complex. Reaction rates were inhibited by added
tin bromide, but not by the arylamine product. The inhibition by tin bromide showed that
reversible transmetalation between an aryl halide complex and the tin reagent was occurring.
Subsequent to reversible transmetalation, a rate-determining reductive elimination of
arylamine occurred. Under conditions with a 10-fold excess of tin amide and high phosphine
concentrations, the rate-determining-step became oxidative addition of aryl bromide, and
reactions became first order, rather than zero order, in aryl bromide. The amido intermediate
deduced from kinetic studies appeared to be generated by reacting {Pd[P(o-Tol)3](p-BuC6H4)-
(Br}2 (2b) with lithium arylamides or by deprotonating {Pd[P(o-Tol)3](NHEt2)(p-BuC6H4)-
(Br)} (3b) with MN(SiMe3)2 (M ) K, Li). Both reactions gave yields of arylamine that were
comparable to those of catalytic reactions. Competition and relative rate studies revealed
an equilibrium between aryl halide complexes 2a -c and a tin amide adduct of it. In
competition studies involving an in situ selectivity for reaction of Bu3SnNMe2 or Bu3SnNEt2
with p-t-BuC6H4Br catalyzed by 1, the ratio of N,N-dimethylaniline to N,N-diethylaniline
was 2.9. However, kinetic measurements of individual reactions showed that Bu3SnNMe2
reacted only 1.4 times faster than Bu3SnNEt2, consistent with a reversible equilibrium
involving tin amide binding to the catalyst, similar to that resulting from substrate binding
pree¨quilibria in enzyme systems.
In tr od u ction
organic molecules, catalytic organometallic reaction
chemistry that leads to the formation of carbon-
heteroatom bonds is less common than that forming
carbon-carbon and carbon-hydrogen bonds. Moreover,
the construction of C-N bonds in amines is particularly
rare.2-11 In large part, routes to the necessary reactive
intermediates for such catalysis and the fundamental
A principal goal of organometallic chemistry is the
catalytic production of organic molecules by exploiting
the distinct reaction chemistry of organic ligands co-
valently bound to transition metals. Most organome-
tallic chemistry has focused on complexes with covalent
metal-carbon or metal-hydrogen bonds. The platinum
group transition metals, in particular palladium and
rhodium, have been workhorse elements in many com-
mercialized catalytic processes that include hydrogena-
tions, hydroformylations, acetic acid production, and
other C-C and C-H bond forming processes.1 Al-
though carbon-oxygen, carbon-nitrogen, or carbon-
sulfur bonds are found in the majority of important
(2) One reaction of this type is the hydroamination of alkenes and
alkynes. For leading references see the following nine citations:
(3) Catalysis with a low valent arylamido complex has been reported,
although turnover numbers are not high: Casalnuovo, A. L.; Calabrese,
J . C.; Milstein, D. J . Am. Chem. Soc. 1988, 110, 6738-6744 and
references therein.
(4) Gagne´, M. R.; Marks, T. J . J . Am. Chem. Soc. 1989, 111, 4108-
4109.
(5) Gagne´, M. R.; Nolan, S. P.; Marks, T. J . Organometallics 1990,
9, 1716-1718.
(6) Brunet, J .; Commenges, G.; Neibecker, D.; Philippot, K. J .
Organomet. Chem. 1994, 469, 221.
X Abstract published in Advance ACS Abstracts, May 15, 1996.
(1) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, 1987.
(7) Pez, G. P.; Galle, J . E. Pure Appl. Chem. 1985, 57, 1917-1926.
(8) Hegedus, L. S.; Allen, G. F.; Bozell, J . J .; Waterman, E. L. J .
Am. Chem. Soc. 1978, 100, 5800.
S0276-7333(96)00188-4 CCC: $12.00 © 1996 American Chemical Society

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2795
reagents are ill-defined.29,30 Further, the copper medi-
ated processes have eluded any clear mechanistic
conclusions.30 Although Kosugi reported over 10 years
ago a palladium-catalyzed amination of aryl halides, the
reaction is relatively unexplored both mechanistically
and synthetically. Recently our group15,23,25,31-33 and
Buchwald’s group22,26 have been working to determine
the synthetic scope of the palladium-catalyzed formation
of arylamines from aryl halides, to develop the simplest
coupling procedures, and to characterize intermediates
in the coupling reaction. It seemed likely to us that the
palladium chemistry would be more receptive than the
copper chemistry to mechanistic analysis and that a
well-understood coupling process would allow for ratio-
nal modification of reaction conditions, development of
new catalysts, and design of substrates that would
participate in the aryl halide aminations. The tin amide
chemistry seemed likely to be most receptive to mecha-
nistic analysis, and the mild tin reagents present some
synthetic advantages, particularly with base-sensitive
systems. Perhaps most important from a mechanistic
standpoint, it seemed likely that a palladium amido
complex was the key intermediate on the reaction
pathway. If this proved to be the case, then its reactiv-
ity would help clear a path for further catalytic applica-
tions of late metal amides and alkoxides.
Sch em e 1
reactions required of such intermediates are poorly
developed. Specifically, synthetic routes to reactive
amido (M-NRR′) and alkoxo (M-OR) complexes have
attracted significant attention only recently, and late
transition metal amido and alkoxo complexes with
â-hydrogens typically decompose rapidly to metal hy-
drides.12,13 As a result, no dialkylamido complexes of
platinum-group metals have been isolated or even
observed,14 and only a few monoalkyl complexes15-19
have been isolated despite the generality of the amido
group.
The addition of tin amides to aryl halides in the
presence of a hindered Pd(0) catalyst forms arylamines
and tin halides by a process that is general for a variety
of dialkyl or aryl amides and aryl halides (Scheme
1).20-26 Turnover numbers can be greater than 100, and
selectivity for arylamines can exceed 95%. This “hetero-
cross-coupling” chemistry is a heteroatom analog of
Stille couplings that form C-C bonds from aryl, vinyl,
or acetylenic tin reagents and sp2-hybridized organic
halides.27,28 The use of the Stille reaction is widespread
in the preparation of natural products and novel organic
materials.
The accepted mechanism for Stille carbon-carbon
bond-forming coupling reactions involves oxidative ad-
dition of an organic halide to a Pd(0) complex, followed
by transfer of a group from an organotin reagent to the
palladium aryl halide complex and reductive elimination
of the organic product to regenerate Pd(0).27,28 Kinetic
studies and stoichiometric chemistry of organopalla-
dium complexes have led to the acceptance of these
general mechanistic features.34
Formation of arylamines from aryl halides has been
a difficult transformation. Ullman chemistry, the reac-
tion of copper amides with aryl halides, has been used
for this transformation in the past, but yields are
variable, high temperatures are necessary, inconvenient
and toxic solvents are required, and the active copper
A rough mechanism for the formation of arylamines
(Scheme 1) that would parallel the accepted mechanism
for Stille couplings requires an amido intermediate and
a catalytic cycle that would be comprised of a number
of unusual features and reactions that are rare or
unprecedented in stoichiometric chemistry. For ex-
ample, an amido intermediate would have to undergo
reductive elimination of arylamine, a reaction with little
precedent, faster than â-hydrogen elimination to yield
imine, a reaction which has been thought to be rapid
for alkylamido complexes with â-hydrogens.12,35 Fur-
ther, formation of a transition metal amide from a tin
amide is unknown and has been shown by our group to
be unfavorable thermodynamically.15
(9) Walsh, P. J .; Baranger, A. M.; Bergman, R. G. J . Am. Chem.
Soc. 1992, 114, 1708.
(10) Baranger, A. M.; Walsh, P. J .; Bergman, R. G. J . Am. Chem.
Soc. 1993, 115, 2753.
(11) McGrane, P. L.; J ensen, M.; Livinghouse, T. J . Am. Chem. Soc.
1992, 114, 5459.
(12) Bryndza, H.; Tam, W. Chem. Rev. 1988, 88, 1163-1188.
(13) Fryzuk, M. D.; Montgomery, C. D. Coord. Chem. Rev. 1989, 95,
1-40.
(14) Examples of a nickel dimethylamido complex were prepared
over 20 years ago: Klein, H. F.; Karsch, H. H. Chem. Ber. 1973, 106,
2438-2454.
(15) Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc. 1995, 117, 4708.
(16) Rahim, M.; Bushweller, C. H.; Ahmed, K. J . Organometallics
1994, 13, 4952-4958.
We have employed a combination of kinetic studies,
stoichiometric reactivity, and independent synthesis of
reaction intermediates to demonstrate that the novel
organometallic chemistry outlined roughly in Scheme
1 does occur. Herein, we identify specifically the
complexes that lie on the catalytic pathway, the primary
reactions that are kinetically important, the steps that
are reversible, and the steps that form product irrevers-
ibly. All of our data indicates that a three-coordinate
(17) Michelman, R. I.; Ball, G. E.; Bergman, R. G.; Andersen, R. A.
Organometallics 1994, 13, 869-881.
(18) Sarneski, J . E.; McPhail, A. T.; Onan, K. D.; Erickson, L. E.;
Reilley, C. N. J . Am. Chem. Soc. 1977, 99, 7376-7378.
(19) A dimeric alkyl, benzyl amide has been prepared: Fryzuk, M.
D.; Piers, W. E. Organometallics 1990, 9, 986-98.
(20) Kosugi, M.; Kameyama, M.; Sano, H.; Migita, T. Nippon Kagaku
Kaishi 1985, 3, 547-551.
(21) Kosugi, M.; Kameyama, M.; Migita, T. Chem. Lett. 1983, 927-
928.
(22) Guram, A. S.; Buchwald, S. L. J . Am. Chem. Soc. 1994, 116,
7901.
(23) Paul, F.; Patt, J .; Hartwig, J . F. J . Am. Chem. Soc. 1994, 116,
5969.
(24) The two following references report aryl halide amination
methods methods without tin reagents.
(25) A tin free system has also been developed. Louie, J .; Hartwig,
J . F. Tetrahedron Lett. 1995, 36, 3609.
(26) Guram, A. S.; Rennels, R. A.; Buchwald, S. L. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1348-1350.
(27) Stille, J . K. Pure Appl. Chem. 1985, 57, 1771-1780.
(28) Stille, J . K. Angew. Chem. Int. Ed. Engl. 1986, 25, 508-524.
(29) Lindley, J . Tetrahedron 1984, 40, 1433-1456.
(30) Paine, A. J . J . Am. Chem. Soc. 1987, 109, 1496-1502.
(31) Hartwig, J . F.; Paul, F. J . Am. Chem. Soc. 1995, 117, 5373.
(32) Louie, J .; Hartwig, J . F. J . Am. Chem. Soc. 1995, 117, 11598.
(33) Paul, F.; Patt, J .; Hartwig, J . F. Organometallics 1995, 14, 3030.
(34) A recent review of these data is the following: Brown, J . M.;
Cooley, N. A. Chem. Rev. 1988, 88, 1031.
(35) Bryndza, H. E.; Calabrese, J . C.; Marsi, M.; Roe, D. C.; W., T.;
Bercaw, J . E. J . Am. Chem. Soc. 1986, 108, 4805-4814.

2796 Organometallics, Vol. 15, No. 12, 1996
Louie et al.
Pd(II) amido complex is formed by transmetalation
between an aryl halide complex and a tin amide and
that this intermediate forms the amine C-N bond by
carbon-heteroatom bond-forming reductive elimination.
These data provide the first firm identification of a
catalytically active, platinum-group alkylamido complex
and reveal its unprecedented selectivity for carbon-
heteroatom bond formation. As a result of this selectiv-
ity, this chemistry produces arylamines during a prac-
tical catalytic nucleophilic aromatic substitution reaction.
by Kosugi, since reduction of these Pd(II) complexes
occurs on a time scale similar to the actual catalytic
amination.
Simple 31P NMR spectroscopic monitoring of reactions
catalyzed by amine complexes 3a or by 1 or 2a in the
presence of amine showed that complex 3a was the
resting state of the palladium. The clear identification
of this species as the resting state led us to conduct
detailed kinetic studies with this catalyst. Potential Pd-
(0) intermediates or Pd(II) amido complexes were not
observed in the reaction medium by spectroscopic
techniques. Amine complexes related to 3a have previ-
ously been characterized fully by spectroscopic and
microanalytical techniques.33 Both 1H and 31P NMR
spectra of reaction solutions showed 3a to be the only
palladium complex present in observable quantities. We
have reported in stoichiometric studies that this amine
complex reacts with tin amides to give Pd(0) and
arylamines.32 The bulky tin amide did not displace the
secondary amine of catalyst 3a to give a tin amide
complex as an observable species, although such a
displacement may be part of the reaction pathway.
Identification of the resting state of reactions catalyzed
by 1 or 2a was more complicated. 31P NMR spectra of
the reaction solutions showed a broad resonance at 29
ppm similar to that of isolated 2b. However, as will be
described below, further studies showed the resting
state to be an equilibrium mixture of 2b and a tin amide
adduct of it, which presumably has a structure similar
to amine complex 3a .
Kin etic Evid en ce for a P a lla d iu m -Am id o In ter -
m ed ia te. A mechanism involving rate determining
reaction of 3a and the tin reagent would provide first-
order rate behavior in tin amide, zero-order behavior
in phosphine, zero-order behavior in aryl halide, first-
order behavior in palladium catalyst, and inverse first-
order behavior in amine. Similar rate behavior would
be expected for reactions catalyzed by 2a , except that
half-order behavior in the palladium catalyst would be
observed if the tin amide reacted with the resulting
monomeric aryl halide complex formed from reversible
cleavage of dimeric 2a . Reactions catalyzed by 1 would
show the same characteristics as those catalyzed by 2a ,
since oxidative addition of aryl bromides to 1 is rapid,
relative to subsequent steps, under normal conditions.
Importantly, rate-determining transmetalation would
prevent our spectroscopic or kinetic detection of the
amido intermediate in the catalytic system.
Resu lts
Our initial mechanistic studies of the catalytic process
involved determination of the reaction order in aryl
halide, tin amide, and palladium catalyst. We con-
ducted our reactions such that turnover numbers were
limited to five. These conditions prevented measurable
catalyst decomposition, as determined by integration of
a
31P NMR spectrum of the reaction solution vs a
phosphine standard contained in a capillary. Reactions
were monitored by the decay of tin amide reagent whose
concentration was conveniently determined by 1H NMR
spectroscopy while heating the sample in the probe.
Quantities of the other reagents were chosen such that
the concentrations of all reagents except that of the tin
amide remained virtually constant. A typical catalytic
solution for kinetic analysis contained 3-5 equiv of
tributyltin dimethylamide, 5-10 equiv of tributyltin
bromide, 10-30 equiv of p-t-BuC6H4Br, and at least 5
equiv of added phosphine for every 1 equiv of catalyst.
Catalyst concentrations were on the order of 0.01 M,
and reaction temperatures ranged from 60 to 85 °C
depending on the catalyst used (see Experimental
Section for details).
Three palladium catalysts were employed for our
mechanistic studies. One complex was the Pd(0) species
{Pd[P(o-Tol)3]2} (1), whose synthesis, spectroscopic char-
acterization, and X-ray structure has been reported by
our group previously.33 A second complex we employed
was the dimeric {Pd[P(o-Tol)3](p-MeC6H4)(Br)}2 (2a ),
resulting from aryl bromide oxidative addition reactions
to 1 (eq 1). Compound 2a also catalyzes the amination
It was expected, but important to confirm, that the
overall process was irreversible. Catalytic reactions
between Bu3SnNEt2 and t-BuC6H4Br in the presence
of excess N,N′-dimethylaniline showed no formation of
p-t-BuC6H4NMe2. Moreover, reactions between Bu3-
SnNMe2 and p-CH3C6H4Br monitored in the presence
and absence of added N,N′-dimethyltoluidine product
showed the same conversions at various reaction times.
Although the overall reaction was irreversible, the
portion of the catalytic cycle that formed tin bromide
as product proved to be reversible. In contrast to
addition of the arylamine product to reaction solutions,
addition of >4 equiv of tributyltin bromide to the
reaction solution led to a dramatic decrease in reaction
rates. In fact, as can be seen by comparing A and B of
Figure 1, addition of an excess of tin bromide to initial
reaction solutions provided linear first-order plots.
As presented in the Discussion, the curvature of the
plot in Figure 1A is due to product inhibition. The
reaction, and its synthesis, spectroscopic characteriza-
tion, and X-ray structure have also been reported by our
group.33 This aryl halide complex undergoes reversible
dimer cleavage before reaction with conventional aryl-
tin reagents such as phenyl trimethyltin.32 Addition-
ally, this dimeric species reacts with primary and
secondary amines as in eq 2 to yield monometallic amine
complexes, {Pd[P(o-Tol)3](NHR2)(Ar)(Br)} (3a , Ar )
p-MeC6H4, R ) Me; 3b, Ar ) p-BuC6H4, R ) Et).33
These amine complexes also catalyze the coupling
between aryl halides and tin amides and were the third
type of complex used in the catalytic studies. We have
employed these three types of compounds rather than
{Pd[P(o-Tol)3]2Cl2} (4), the catalyst originally reported

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2797
F igu r e 1. Results of monitoring the decay of tin amide
during reaction of Bu3SnNMe2 with p-CH3C6H4Br cata-
lyzed by 3a at 75 °C (graph A) or identical reaction
conditions, but containing 4 equiv of added Bu3SnBr and
conducted at 85 °C (graph B).
addition of excess tin bromide to the reaction solutions
ensures little change in tin bromide concentrations
throughout the reactions, and first-order decay of tin
amide is observed.
F igu r e 2. Determination of the order in 3a for the reaction
of p-MeC6H4Br with Bu3SnNMe2 catalyzed by 3a .
The dramatic rate decrease in the presence of tin
bromide was not due to catalyst decomposition, as initial
rates also showed a strong inhibition upon addition of
tin bromide. Further, reaction mixtures containing
small concentrations of added tin bromide did not show
catalyst decomposition for up to 12 h at room temper-
ature or a rate of catalyst decomposition that signifi-
cantly exceeded that of reactions containing no added
tin bromide. As suggested by these results, added tin
bromide did not affect yields of the oxidative addition
of p-CH3C6H4Br by 1. For example, compound 1 reacted
with p-CH3C6H4Br to form 2a in 85-90% yield in the
presence of 5 equiv of tin bromide. It should be noted,
however, that large concentrations of tin bromide did
lead to slow decomposition of the catalyst, and tribu-
tyltin bromide reacted at 70 °C with aryl halide complex
2a to give a black Pd precipitate in the absence of added
aryl bromide. This slow tin bromide induced decompo-
sition of catalyst led to low arylamine yields at artifi-
cially high concentrations of tin bromide and, thus,
prevented our determining an accurate reaction order
in tin bromide. At low concentrations of tin bromide
(0.050-0.15 M), however, reactions did show a rough
inverse first-order behavior in added tin bromide,
although the lower concentrations in this range did not
provide strictly pseudo-first-order reaction conditions
preventing quantitative information.
As a result of this rate inhibition by tin bromide and
the need for excess tin bromide to obtain linear first-
order plots, quantitative rate measurements were con-
ducted in the presence of 5 equiv of added tin bromide
per 1 equiv of tin amide. Under these conditions,
reactions between p-CH3C6H4Br and Bu3SnNMe2 in the
presence of catalytic amounts of 3a stabilized by added
P(o-Tol)3 were zero order in free aryl bromide from 0.21
to 1.01 M. Reaction rates were also independent of the
phosphine concentration from 0.11 to 0.57 M. Varying
the concentrations of these reagents provided first order
rate constants that were all within 10% of 2.0 × 10-4
s-1. Figure 2 shows the results of varying the catalyst
concentration from 0.006 to 0.030 M. Catalysis involv-
ing a dimeric palladium complex would show second-
order behavior in 3a ; our results indicate first-order,
rather than second-order, behavior in catalyst and show
that the catalysis involves monomeric palladium com-
plexes. Rate constants for reactions catalyzed by 3a
were measured in the presence of 0.051 and 0.27 M of
added HNMe2, and an inverse first-order dependence
of rate on amine concentration was observed as shown
in Figure 3.
F igu r e 3. Determination of the order in HNMe2 for the
reaction of p-MeC6H4Br with Bu3SnNMe2 catalyzed by 3a .
A complete kinetic analysis of reactions between p-t-
BuC6H4Br and Bu3SnNMe2 catalyzed by 1 or 2a was
not conducted in the presence of added tin bromide.
However, qualitative data obtained in the absence of
added tin bromide by comparing conversions at a series
of reaction times over at least 3 half-lives showed that
these reactions had the same overall features as those
catalyzed by 3a sfirst order in tin amide, zero order
in aryl bromide, and zero order in added phosphine.
Reaction rates were unaffected by free aryl bromide
concentration (0.10-0.40 M) and were independent of
the concentration of added phosphine (0.05-0.21 M).
Again, adding tin bromide led to inhibition of the
reactions. Monitoring the decay of tin amide during
experiments that were conducted with added tin bro-
mide gave linear first-order plots, confirming the first-
order behavior in tin amide. It was experimentally
difficult to distinguish half from first-order behavior in
palladium catalyst when starting from the dimer 2a .
However, as presented in the Discussion, the clear
presence of a three-coordinate, monomeric intermediate
in the reactions of 3a strongly suggests that the
catalytically active form of 2a is mononuclear.
Electr on ic Effects. In order to probe for direct
attack of the tin amide at the palladium-bound aryl
group, we conducted the catalytic reactions with several
aryl bromides possessing different para substituents.
Rate constants were compared for reactions of Bu3-
SnNMe2 and p-XC6H4Br (X ) t-Bu, CF3, F, NMe2, OMe)
catalyzed by 1. Our results given in Table 1 showed
that the rates were slightly faster as the electron density

2798 Organometallics, Vol. 15, No. 12, 1996
Louie et al.
Ta ble 1. Ra te Con sta n ts for Cou p lin g Rea ction s
of a palladium amide had not occurred. Addition of 1
equiv of tin amide left a large, sharp tert-butyl resonance
for 2b, but also formed a small, broad resonance at the
same chemical shift as the resonance observed upon
addition of 10 equiv of tin amide. Unfortunately,
accurate integrations could not be obtained, due to large
nearby butyl resonances of the tin amide. Incremen-
tally increasing amounts of tin amide, however, did lead
to a clear increase in the broad adduct resonance and a
disappearance of the sharp tert-butyl resonance of the
aryl halide complex 2b. Cooling of the samples below
room temperature led to complex 1H and 31P NMR
spectra, as expected, since even spectra for analytically
pure dimeric aryl halides are extremely complex at low
temperatures.33 Unfortunately, one cannot draw firm
conclusions from 31P NMR spectra of samples containing
2b and Bu3SnNMe2 at ambient temperatures, since only
slight differences in line shape were observed. Although
the spectroscopic data is not straightforward, it is clear
that 2a and 2b reversibly react with tin amide without
forming tin bromide. The simplest explanation for this
reversible reactivity is shown in eq 3 and involves the
w ith Differ en t Ar yl Br om id es p-R-C6H4Br
R
σ
rate const (s-1 × 10-4
)
relative rates
CF3
F
t-Bu
MeO
Me2N
0.54
0.06
-0.19
-0.28
-0.83
9.0
6.5
4.8
2.9
2.4
3.8
2.7
2.2
1.2
1.0
at the aryl ring was decreased. However, reaction rates
for the two extreme cases, p-CF3- and p-NMe2-substi-
tuted aryl halides, were different by a factor of only 3.8.
Effect of High Con cen tr a tion s of P h osp h in e.
Reactions conducted with high concentrations of tin
amide and phosphine showed a different resting state
than those with low concentrations (below 40 mM) of
added phosphine and similar amounts of tin amide to
aryl halide. Specifically, reactions conducted with either
1 or 2a , a 10:1 ratio of tin amide to aryl bromide, and
concentrations of free phosphine near 100 mM showed
only resonances for free phosphine (-29 ppm) and L2-
Pd0 complex 1 (-6 ppm). Moreover, linear first-order
plots for the decay of aryl bromide were obtained,
contrasting the zero-order behavior observed at lower
concentrations of phosphine and tin amide. Clearly the
rate-determining-step of the catalytic cycle was shifted
from a combination of transmetalation and reductive
elimination to aryl halide oxidative addition.
formation of 8, a tin amide adduct of 2a and 2b that
would have a structure similar to those of amine
complexes 3. The similarity of the 31P NMR chemical
shifts to those of amine complexes 3 are consistent with
this conclusion.
Reactions run without added phosphine gave lower
yields of arylamine than those run in the presence of
added phosphine, and formation of black palladium(0)
precipitate was observed. Addition of 5 equiv of P(o-
Tol)3 per 1 equiv of catalyst kept reaction mixtures
homogeneous, thereby increasing turnover numbers and
product yields. Increasing the phosphine concentration
from 5-15 equiv induced a slight increase in the rate,
due to further prevention of catalyst decomposition,
particularly when conducting reactions with Pd(0)
complex 1 or the dihalo complexes Pd[P(o-Tol)3]2Br2 (5)
and 4. However, addition of 20 equiv of phosphine led
to a measurable decrease in rate. Under these condi-
tions, a 31P NMR signal was observed at -6 ppm and
first-order rate behavior in aryl bromide and palladium
catalyst were obtained, consistent with rate-determining
oxidative addition and a change in resting state to 1.
Reactions involving reversible formation of a complex
between substrate and catalyst often provide product
ratios from reactions containing more than one sub-
strate that are different from the ratio of rate constants
obtained from two reactions, each involving one of the
substrates. This situation is common for enzyme sys-
tems and is termed competitive inhibition.37 In our
case, the apparent equilibrium between free and sub-
strate-bound catalyst would suggest that competitive
inhibition is likely to be observed and that different
relative rates would be obtained from competition
experiments and from individual rate measurements.
Quantitative rate measurements for reactions of tin
amides catalyzed by 1 showed that rate constants for
Bu3SnNMe2 were 1.4 ( 0.1 times faster than those of
Bu3SnNEt2. However, the selectivity for reactions of
p-BuC6H4Br with a mixture of these two reagents was
2.9 ( 0.4.38 The differences between these values
suggest that reversible binding of substrate occurs and
is consistent with the apparent equilibrium between free
dimeric 2 and a tin amide-bound species.
F or m a tion of a Tin Am id e Ad d u ct. Determination
of the resting state of the palladium for reactions
catalyzed by 1 or 2b did not prove to be as straightfor-
ward as identifying the resting state for reactions
catalyzed by 3a . A broad resonance at 29 ppm was
observed that could be attributed to the dimeric aryl
halide complexes. It should be noted that both aryl
halide complex 2a and its amine adducts 3a and 3b
have essentially identical 31P NMR chemical shifts.33
In stoichiometric studies, addition of tin amide to the
dimeric aryl halide complexes led to little change in the
31P NMR chemical shift, but 1H NMR spectra showed
distinct differences indicating that reaction had oc-
curred. For example, upon addition of 10 equiv of Bu3-
SnNMe2 to the p-t-BuC6H4 derivative of 2a (2b, {Pd-
[P(o-Tol)3](p-t-BuC6H4)(Br)}2) a new, broad tert-butyl
resonance was observed slightly downfield of that for
2b.36 No tin bromide was formed, demonstrating that
simple substitution of amide for halide and generation
In d ep en d en t Gen er a tion of th e Am id o Com p lex.
Rea ction of {P d [P (o-tolyl)3](p -Bu C6H4)(Br )}2 (2c)
w ith Lith iu m Am id es. We sought to support our
kinetic studies by synthetic studies aimed at indepen-
dently generating the tri-o-tolylphosphine-ligated pal-
ladium amido complexes. Initially, we reacted aryl
(37) Segel, I. H. In Enzyme Kinetics; J ohn Wiley and Sons, Inc.: New
York, 1975; pp 113-118.
(38) Buchwald and Guram reported a 2:1 selectivity for formation
of N,N-dimethylanilines over N,N-diethylanilines as a competition
study. Their reactions were run in the presence of 1.4 equiv of both
the diethyl- and dimethylamido tin reagents per equivalent of aryl
bromide. In order to obtain the accurate value that our comparison
requires, a constant concentration of the two tin amides must be
ensured. With the 10-fold excess of each tin amide in our experiments,
a 3:1 selectivity was observed.
(36) 1H NMR spectra were different from the amine complex, ruling
out formation of 3 by an amine impurity in the tin amide.

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2799
Sch em e 2
halide complex 2c with lithium amides, as shown in eq
4. Addition of lithium amides to a slurry of 2c in THF
solvent caused dissolution of the aryl bromide complex
and produced an orange or red solution accompanied
by a black palladium precipitate. As determined by a
combination of 31P NMR spectroscopy and GC, the
solution contained L2Pd complex 1, free phosphine, and
arylamine. However, the formation of arylamine oc-
curred in yields varying from only 35% to 85% as
determined by GC analysis against a naphthalene
internal standard. These yields were well below those
observed during catalytic experiments employing tin
amides.
Control reactions demonstrated that lithium amides
reacted to a small extent with p-BuC6H4Br under the
same conditions, giving two isomeric arylamines in 22%
and 17% yields as well as arene in 23% yield. These
product ratios were different from those obtained from
the more selective reaction with 2c. Most importantly,
only one arylamine isomer was formed from addition
of lithium amides to 2c.
of arylamine and [P(o-Tol)3]2Pd0 products in greater
than 90% yield and regenerated Pd(0). As communi-
cated earlier, this reaction became the basis for the
catalytic amination of aryl halides in the absence of
tin.25
Addition of the lithium silylamide at low temperature
allowed spectroscopic detection of the anionic amido
complex, M{Pd[P(o-Tol)3](p-BuC6H4)(Br)(NEt2)} (M )
Li, K) (7). The starting amine complex, which displayed
a singlet 31P NMR spectroscopic resonance at δ 29 was
converted at -70 °C to a new species displaying a
resonance at δ 25. No N-H resonance for a coordinated
1
amine was observed in the low temperature H NMR
spectra obtained after addition of the silylamide, and
HN(SiMe3)2 was observed. These results imply that a
palladium amido complex was formed by the low tem-
perature deprotonation.
However, further experiments showed that this amido
complex was not a neutral species. The use of KN-
(SiMe3)2, rather than the lithium reagent, led to conver-
sion of the amine complex to a set of compounds
displaying resonances between δ 20 and δ 24 at -70
°C, rather than the single resonance at δ 25. The
presence of several resonances was not simply due to
the irreversible formation of multiple reaction products,
as these compounds underwent interconversion on the
NMR time scale. The set of signals observed at -70 °C
coalesced at -40 °C to a single peak at δ 22. Warming
of this complex to -20 °C led to formation of arylamine
and 1 after 20 min. The analogous complex formed by
deprotonation by LiN(SiMe3)2 was stable at -20 °C for
at least 1 h, but slowly formed arylamine and L2Pd0
product 1 at 0 °C.
The differences in spectral features and reaction rates
for the amido complexes formed from LiN(SiMe3)2 and
KN(SiMe3)2 strongly suggested that the amido com-
plexes formed at low temperatures are anionic species
7 with interconverting geometries and aggregation
numbers. Warming these complexes led to loss of the
alkali metal bromide, along with formation of arylamine
and Pd(0) without direct detection of a neutral amido
species. However, the formation of a neutral amido
complex by loss of lithium or potassium bromide is the
most reasonable pathway for reaction of 7, as argued
in the discussion.
Variation of the lithium amide concentration, reaction
temperature, or added phosphine concentration had
either a detrimental effect or only a small positive effect
on the yields of the reactions of 2c with lithium amides.
However, the identity of the amide had a measurable
influence. Higher yields were observed with lithium
diethylamide, 6b, than lithium dimethylamide, 6a ,
under the same conditions. Surprisingly, the most
dramatic difference was observed with arylamides.
Addition of lithium anilide, 6e, or lithium diphenyla-
mide, 6f, led to yields of diaryl- or triarylamines that
were competitive or exceeded catalytic reactions (66-
95%) with tin amides. Since the yields of these sto-
ichiometric additions of arylamides were similar to those
observed during catalysis, we concluded that the lower
yields with the less soluble lithium dialkylamides
resulted either from a detrimental effect of the excess
amounts of these highly reducing reagents on the
chemistry of the resulting amido complex or from a low
yield in the generation of the expected amido intermedi-
ate. We, therefore, sought an alternative and milder
method for the independent generation of the dialky-
lamido intermediate.
Dep r oton a tion of th e Am in e Com p lex {P d [P (o-
Tol)3](p -Bu C6H4)(Br )(HNEt2)} (3b). Our methods for
successful independent generation of alkylamido inter-
mediates are shown in Scheme 2. The N-H bonds in
metal amine complexes are more acidic than those in
free amines and are deprotonated irreversibly by lithium
amides.39-41 As shown in Scheme 2, reaction of the
soluble, hindered amide LiN(SiMe3)2 with amine com-
plex 3b in the presence of phosphine led to formation
Sa lt Effects. Due to low solubility in aromatic
solvents, added lithium and potassium salts had no
influence on the catalytic reaction. However, added
salts did have an effect on reactions conducted in THF
solvent. Amination reactions catalyzed by 1 in the
presence of LiBr or LiCl resulted in less than 100%
conversion of aryl halide (88%) to give arylamine (53%)
and arene products including various biphenyls.42-46
LiBr changed the color of homogeneous reaction mix-
tures from yellow to orange.
(39) Park, S.; Roundhill, D. M.; Rheingold, A. L. Inorg. Chem. 1987,
26, 3972-3974.
(40) Park, S.; Rheingold, A. L.; Roundhill, D. M. Organometallics
1991, 10, 615-623.
(41) J oslin, F. L.; J ohnson, M. P.; Mague, J . T.; Roundhill, D. M.
Organometallics 1991, 10, 2781-2794.

2800 Organometallics, Vol. 15, No. 12, 1996
Louie et al.
Sch em e 3
The presence of dissolved KF gave much lower yields
of arylamines and slightly higher yields of hydrogenoly-
sis product. Only 35% arylamine product was produced
when using [P(o-tolyl)3]2PdBr2 (5), and conversion of aryl
halide was not complete (61%). Conversely, a 1.5 h
reaction run with 1 in THF gave full conversion of the
aryl bromide. Yet, only 68% arylamine was produced.
All reactions conducted in the presence of KF involved
either catalyst 5 or 1, and both formed a black Pd(0)
precipitate. Thus, added halide salts generally had little
effect on reactions run in aromatic solvents but had a
detrimental effect on reactions in THF, especially when
the Pd(II) precursor 5 was employed.
transmetalation, and reductive elimination appears to
be followed, but we have now obtained a variety of
kinetic and synthetic data that clearly identify com-
plexes lying on the reaction pathway and delineate the
relative rates of the individual steps and the reversiblity
or irreversiblity of them. These important modifications
of Scheme 1 are included in Scheme 3.
Copper halides had a far more detrimental effect on
the amination chemistry than lithium chloride or potas-
sium flouride. CuCl, CuBr, and CuI all promoted the
formation of black Pd(0) precipitates in either THF or
toluene. Although conversion of the aryl halide in THF
was complete in the presence of CuCl and CuBr, the 2
h reaction at 75 °C gave only 5-7% arylamine product.
Arene was the major organic product in this case.
Similarly, 92-98% conversion of aryl halide occurred
in the presence of CuCl or CuBr in toluene solvent, but
yielded only 3-12% arylamine. Addition of CuI led to
low turnover numbers in either solvent. Only 2-3 equiv
of aryl halide per equivalent of catalyst reacted in THF
solvent and only 10-15 equiv of aryl halide reacted in
toluene. Some conversion of the tin amide to the
corresponding tin halide occurred in a separate reaction
between Bu3SnNMe2 with CuCl and CuI as determined
by GC/MS analysis of the reaction solution.
Solven t Effects. A comparison of rates measured
in THF and toluene solvents showed subtle differences
between the conditions employing high aryl halide
concentrations and conditions employing high tin amide
concentrations. Conversion of the aryl halide was
complete and gave yields greater than 85% in both
solvents. However, qualitative comparison of reaction
rates showed rates in THF solvent that were roughly
twice as fast as the same reactions conducted in toluene.
These qualitative data were obtained by measuring tin
amide conversions at various times during reactions
catalyzed by 1 and containing excess aryl halide.
In contrast, reactions with high tin amide concentra-
tions and limiting aryl halide concentrations conducted
in THF were slower than those in toluene by a factor of
three. Yields of arylamine (93%) were again similar.
On the other hand, more polar solvents such as PhNO2,
NMP, and CH3CN apparently led to catalyst decomposi-
tion before complete conversion of aryl bromide. Thus,
dramatically lower yields (10-17%) of arylamine prod-
uct were obtained in these media, low conversions were
observed, and reaction rates were greatly reduced.
Restin g Sta te of th e Ca ta lyst. Rea ction s Ca ta -
lyzed by Am in e Com p lexes. Monitoring catalytic
1
reactions by 31P and H NMR spectroscopy showed that
the resting state was clearly the amine complex 3a . A
1
combination of H and 31P NMR spectroscopy conclu-
sively showed that the tin amides did not compete with
amine for a coordination site.
Rea ction s Ca ta lyzed by 1 or 2. It was obvious
from 31P NMR spectroscopy of reaction solutions involv-
ing 1 or complexes 2 as catalyst that their resting state
was not L2Pd0 under conditions where the tin amide and
aryl halide complexes were present in equal concentra-
tions or when the aryl halide substrate was present in
excess. Initially, it appeared that the resting state of
the catalyst was simply the dimeric aryl halide com-
plexes 2. However, 1H NMR spectroscopy suggested
that these aryl halide complexes equilibrated with a tin
amide adduct 8. This conclusion was supported by
conversion of the sharp tert-butyl resonance of 2b to a
separate broad resonance at high concentrations of tin
amide, the observation of both signals at room temper-
ature when a 1:1 ratio of 2b and tin amide was present,
and the lack of formation of tributyl tin bromide during
these observations. Thus, we tentatively concluded that
the dimeric aryl halide complexes were reversibly
cleaved during the catalytic reactions and formed com-
plexes with monometallic structures similar to those of
the isolated and fully characterized amine adducts 3.
Conclusions from spectroscopic data were tentative
because clear spectroscopic data is difficult to obtain on
these tri-o-tolylphosphine complexes. However, inde-
pendent evidence for reversible formation of a catalyst-
substrate complex was obtained by a comparison of
relative rate data under different conditions. Specifi-
cally, a 2.9:1 ratio of two arylamine products was
observed during competition studies, and this selectivity
contrasted the 1.4:1 ratio of rate constants obtained
from individual reactions with only one of the two tin
amides. Such differences in relative rates are signa-
tures of competitive inhibition and signal reversible
binding of substrate to catalyst.
Discu ssion
All of our data are consistent with a mechanism for
the coupling chemistry that involves a three-coordinate
palladium amido complex {Pd[P(o-Tol)3](Ar)(NR2)} (9).
A general mechanism involving oxidative addition,
(42) A variety of biaryls are formed under these conditions, presum-
ably by exchange of palladium- and phosphorus-bound aryl groups
similar to that in the following four leading references.
(43) Baran˜ano, D.; Hartwig, J . F. J . Am. Chem. Soc. 1995, 117, 2937.
(44) Morita, D. K.; Stille, J . K.; Norton, J . R. J . Am. Chem. Soc. 1995,
117, 8576.
Effect of P h osp h in e Con cen tr a tion a n d Su b-
st r a t e R a t ios on t h e R est in g St a t e. Not only do
the tin amide adducts such as 8 lie close in energy to
free tin amide and dimeric complexes 2 but activation
energies for oxidative addition of aryl bromide and
reductive elimination of arylamine are close enough that
(45) Wallow, T. I.; Novak, B. M. J . Org. Chem. 1994, 59, 5034.
(46) Kong, K.-C.; Cheng, C.-H. J . Am. Chem. Soc. 1991, 113, 6313.

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2801
different concentrations of reagents can change the rate-
determining-step. Under typical conditions of a syn-
thetic catalytic experiment with this system, a combi-
nation of transmetalation and reductive elimination of
arylamine determined the rates of reactions. However,
at high phosphine concentrations and in the presence
of excess tin amide, the resting state of the palladium
clearly became L2Pd complex 1 rather than the dimeric
complexes 2 and their tin amide adducts. Under these
conditions, oxidative addition rather than reductive
elimination was rate limiting, and first-order behavior
in aryl halide was observed. We have previously
demonstrated that the oxidative addition of aryl halides
to 1 has an inverse first-order dependence on phosphine
concentration, indicating that phosphine dissociation
precedes irreversible carbon-halogen bond cleavage.31
As a result of this oxidative addition mechanism, high
concentrations of phosphine retard the step of the
catalysis involving free aryl halide. If the phosphine
concentration is high and the ratio of aryl bromide
concentration to tin amide concentration is low, then
the oxidative addition step becomes slower than the
transmetalation step. Small amounts of added phos-
phine, however, assist in inhibiting irreversible catalyst
decomposition.
Kin etic Evid en ce for a n Am id o In ter m ed ia te:
Rever sible Tr a n sm eta la tion a n d Electr on ic Ef-
fects. Rever sible Tr a n sm eta la tion . The combina-
tion of phosphine dissociation and organic group trans-
fer is typically the rate-determining-step of C-C bond-
forming Stille coupling reactions, and we have shown
that dimer cleavage and aryl group transfer is the rate-
determining-step in reactions of dimeric complexes 2
with aryltin reagents.32 Thus, it seemed likely that an
analogous amido group transfer to a three-coordinate
aryl halide complex formed by either amine dissociation
from 3 or dimer cleavage of 2 would be rate determining.
Our observation of 2b and its apparent tin amide adduct
as the resting state was initially consistent with this
expectation. However, nonlinear first-order plots ob-
tained from monitoring the decay of tin amide during
both catalytic and stoichiometric reactions run without
added tin bromide implied our initial mechanistic
hypothesis was too simplistic.
Indeed, added tin bromide did inhibit reaction rates.
This inhibition was clearly observed when qualitatively
monitoring reaction rates and was also assessed quan-
titatively. For example, a rough inverse first-order
behavior was observed at low concentrations of tin
bromide, comparable to concentrations generated during
catalysis. Perhaps most convincing, the curvature of
first-order plots conducted in the absence of added tin
bromide was not observed in the presence of added tin
bromide as shown in Figure 1. This effect is consistent
with curvature due to product inhibition. The curvature
is eliminated by creating conditions under which the
product concentration remains constant.
One must also consider whether the added tin bro-
mide leads to a change in the identity of the catalyst or
leads to decomposition of the catalyst. Although we did
observe that decomposition of catalyst occurred in
competition with the production of arylamine in the
presence of a 20-30-fold excess of tin bromide for each
tin amide (100-150 times the catalyst concentration),
several pieces of data besides the changes in tin amide
decay curve shape in the presence and absence of tin
amide (Figure 1) are not consistent with decreasing
catalyst concentration leading to the mechanistically
important inhibitory effect at lower concentrations of
added tin bromide. First, the initial rates were slower
in the presence of even low concentrations of tin
bromide. Second, quantitative studies of the stoichio-
metric reactions of monometallic amine complex 3 with
tin amides showed inhibition by tin bromide.32 Finally,
reaction mixtures containing small quantities of tin
bromide did not show significant catalyst decomposition,
as determined by 31P NMR spectroscopy involving a
PPh3 standard contained in a capillary tube, even
though much slower reaction rates were observed. The
absence of significant catalyst decomposition at low
concentrations of tin bromide is consistent with the slow
reactivity of complexes 2 with tributyltin bromide and
the selective reaction of 1 with aryl bromides rather
than tributyltin bromide presented in the Results.
Electr on ic Effects: In itial Reaction of Tin Am ide
a t th e P a lla d iu m Cen ter Ra th er Th a n a t th e Ar yl
Gr ou p . Although these kinetic experiments demon-
strated the presence of an intermediate after reaction
of aryl bromide complexes 2b or amine complex 3a with
tin amide that could be 9, they provided no structural
basis for identifying the proposed amido complex as the
intermediate formed. Specifically, it was possible that
the reversibly formed intermediate was a species that
resulted from attack of the tin amide directly on the
palladium-bound aryl ring as presented in eq 6. This
During recent experiments with triphenylphosphine-
ligated palladium amido complexes, we determined that
transfer of an amido group from palladium to tin had a
negative free energy.15 Therefore, transmetalation of
an amido group from tin to tri-o-tolylphosphine-ligated
palladium almost certainly has a positive free energy.
This information led us to hypothesize that the curva-
ture of the first order plots in Figure 1A resulted from
rate inhibition by tin bromide. Since transfer of the
amido group is thermodynamically disfavored, it would
be reversible if reductive elimination were slower than
reversion to the aryl bromide complex (eq 5). In this
mechanism would account for the reaction orders we
observed since uncatalyzed nucleophilic aromatic sub-
stitution reactions involve analogous disruption of the
aromatic π-system.47 However, catalytic reactions run
with several aryl bromides possessing different p-
substituents suggested otherwise. Direct addition of
nucleophiles to aryl systems substituted with strong
electron donors such as -NMe2 are known to react
case, C-N bond-forming reductive elimination would be
the irreversible step in our modified mechanism involv-
ing a palladium amide. Increasing amounts of tin
bromide product would displace the pree¨quilibrium in
eq 5 toward the starting aryl bromide complex and
would inhibit the overall reaction rate.
(47) March, J . Advanced Organic Chemistry; 3rd ed.; J ohn Wiley
and Sons: New York, 1985.

2802 Organometallics, Vol. 15, No. 12, 1996
Louie et al.
several orders of magnitude slower than those substi-
tuted with strong electron acceptors.48 We found that
rates for aryl halides substituted with -CF3 and -NMe2
groups were different by a factor of only 3.8. Such small
rate differences are inconsistent with transfer of the
amide group from the tin to the aromatic system.
Instead, these data support transfer of the amido group
from the tin reagent to the metal center, creating a
neutral palladium-amido intermediate that would un-
dergo irreversible reductive elimination. The small
acceleration by electron-withdrawing groups is consis-
tent with the electronic effects of sulfide elimination,
another type of carbon-heteroatom bond forming re-
ductive elimination that we have reported.43
Rea ction Or d er s w ith Ca ta lyst 3a a n d Th eir
Rela tion sh ip to Ca ta lysis w ith 1 or 2b. The use of
catalyst 3a in the kinetic studies allowed for clear
determination of the nuclearity and coordination num-
ber of the complex that is involved in the irreversible
step of the catalysis. The first-order rate behavior in
the concentration of 3a shows that the transition state
of the irreversible step involves a mononuclear species.
Further, the lack of dependence on phosphine concen-
tration and inverse first-order dependence on amine
concentration show that the complex contains phosphine
as the only dative ligand. The clear identification of the
resting state as an amine-ligated aryl halide complex
shows that the covalent ligands include an aryl group
and either a halide or an amide. The electronic proper-
ties of the reaction show that the tin amide transfers
its amido group to palladium, rather than the aryl ring,
and forms an amido complex. The reversibility of the
transmetalation step shows that the amido species is
involved in the kinetically important, irreversible step
of the catalysis. The first-order dependence of the
reaction rate on Bu3SnNMe2 shows that the amido
complex is not coordinated to tin amide during its
irreversible reaction. Further, the inhibition of the
reaction rate by free tin bromide shows that the amido
complex is not coordinated to tin bromide during its
irreversible reaction. Thus, the kinetically important,
irreversible step involves the Pd(II) complex {Pd[P(o-
tolyl)3](Ar)(NR2)} (9), an amido complex that contains
one phosphine and one aryl group.
palladium complex as would be formed from cleavage
of dimer 2b. The inverse first-order behavior in amine
concentration and the first order rate behavior in
palladium for reactions catalyzed by 3a , along with the
parallels between the kinetics in the presence and
absence of amine, imply that reactions of dimeric 2b in
the absence of amine involve a mononuclear species in
the kinetically important irreversible step. Thus, reac-
tions catalyzed by 1 and 2b most likely involve irrevers-
ible arylamine reductive elimination from the same
three-coordinate amido complex 9 that forms during
reactions of 3a .
The most concise mechanism for formation of 9 during
reactions involving 1 or 2 is loss of tin bromide from
the tin amide adduct 8. Further, the first-order, rather
than second-order, rate behavior in tin amide for reac-
tions catalyzed by 3a make it unlikely that free Bu3-
SnNMe2 reacts with tin amide adduct 8 to generate an
amido complex. It is not, however, rigorously ruled out
that more than one tin amide is involved in the
formation of 9. The irreversible reaction must involve
a single amido group, however, since reactions involving
3a as resting state are first order in tin amide.
In d ep en d en t Gen er a tion of th e Am id o In ter m e-
d ia te. Stoichiometric, synthetic studies allowed for
independent generation of the proposed amido inter-
mediate and demonstrated the chemical competence of
such a species for arylamine formation. Although the
neutral amido intermediate on the catalytic cycle is too
reactive to observe, the independent synthetic route
allowed us to observe an anionic palladium dialkylamido
complex by low temperature NMR spectroscopy. This
anionic amido complex most reasonably leads to the
same neutral amido intermediate 9, which was deduced
from our kinetic studies. Thus, these synthetic experi-
ments allowed us to evaluate the reaction chemistry of
monomeric palladium dialkylamido reactive intermedi-
ates by generating them through a pathway that is
separate from the catalytic process with tin amides.
The stoichiometric reactions of silylamide base with
amine complexes 3 were likely to be initiated by
deprotonation of the coordinated amine, and consistent
with this expectation the silylamine HN(SiMe3)2 was
formed. Subsequent loss of LiBr or KBr would then give
rise to the same three-coordinate amide 9, whose
presence in the catalytic chemistry was deduced from
kinetic studies (Schemes 2 and 3). The differences in
spectroscopic characteristics and the different rates of
reactivity between compounds generated from potas-
sium and lithium silylamides demonstrated that LiBr
and KBr were maintained in the palladium coordination
sphere of the observed compound. We propose that the
amido complexes observed at low temperatures are the
anionic amido species 7 drawn in Scheme 2 and isomeric
or oligomeric forms of it, and that these compounds are
stable to reductive elimination as a result of their
electron rich character. Loss of LiBr at 0 °C or KBr at
-20 °C would then form the neutral amido intermediate
that would undergo rapid reductive elimination of amine
and form [P(o-Tol)3]2Pd0.
It is not required that the reversible transmetalation
step involve amine dissociation. However, transfer of
carbon and sulfur nucleophiles from tin to palladium
involves three-coordinate monophosphine intermedi-
ates.32 The transfer of amides is, therefore, likely to
involve a similar three-coordinate monophosphine in-
termediate that would form by amine dissociation from
amine complexes such as 3a .
Since reactions catalyzed by Pd(0) complex 1 or aryl
halide complexes 2 show reaction orders in aryl halide,
tin amide, added tin bromide, and added phosphine
that are analogous to those observed for reactions
catalyzed by amine complex 3a , it is most reasonable
that reactions involving 1, 2b, and 3a follow similar
pathways. As shown in eq 7, dissociation of amine from
An alternative pathway could involve reductive elimi-
nation of amine from the anionic amido complex and
subsequent extrusion of LiBr or KBr from the Pd(0)
product. However, this alternative pathway would
involve an unusual reductive elimination from a par-
ticularly electron-rich anionic palladium complex. Fur-
ther, this pathway would involve faster reductive elimi-
3a gives rise to the same three-coordinate monomeric
(48) Bunnett, J . F.; Zahler, R. E. Chem. Rev. 1951, 49, 273-412.

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2803
Sch em e 4
nation from the more ionic potassium palladate than
from the lithium palladate. In other words, the more
electron-rich palladium center would undergo reductive
elimination faster than the more electron-poor one.
Thus, initial formation of neutral amido complex 9 that
undergoes reductive elimination is a more reasonable
pathway, albeit not the only possible one.
Sa lt a n d Solven t Effects. Carbon-carbon bond-
forming coupling reactions are significantly influenced
by the presence of anions capable of acting as covalent
ligands and by solvents with high polarity and Lewis
base character. The beneficial effect of chloride in the
coupling reactions involving organic triflates is, presum-
ably, due to the creation of a palladium halide inter-
mediate that is more stable than the corresponding
triflate.27,49,50 However, the beneficial effect of copper
halides has been shown to result from their ability to
coordinate phosphine and thereby effectively lower the
free phosphine concentration.51 These added salts are
beneficial in polar coordinating solvents which not only
solubilize the added salt but also can act as good ligands
to further stablilize unsaturated intermediates.52 The
amination chemistry showed no beneficial effects of
added halides. No salt effects from added lithium
chloride or potassium fluoride were observed in benzene,
presumably due to the insolubility of them, but some
detrimental effects were observed in THF. Copper
halides radically reduced the production of arylamines
in both THF and toluene.
The effect of employing THF vs toluene solvent
depended on the catalyst resting state. Under normal
catalytic conditions where tin amide concentrations
were greater than or comparable to aryl halide concen-
trations, transmetalation and reductive elimination are
rate determining. Polar and donating solvents are
known to enhance the rate of transmetalation involving
tetraalkyltin reagents,53 and consistent with these data,
reaction rates were slightly faster in THF. However,
solvents that are more polar and coordinating than THF
led to catalyst decomposition. Apparently, intermedi-
ates in the coupling with tin amides decompose more
rapidly in polar solvents than intermediates in C-C
bond-forming Stille chemistry.
Conversely, reactions rates in THF were slower than
those in toluene by a factor of 3 when aryl halide
concentrations were low relative to tin amide and
phosphine concentrations and aryl halide oxidative
addition was rate determining. Aryl halide oxidative
additions to (PPh3)4Pd0 are known to be slightly slower
in THF solvent than in aromatic solvents.54 Thus, the
slower rates in THF are consistent with the previous
data concerning solvent effects on aryl halide oxidative
addition to Pd(0).
â-Hyd r ogen Elim in a tion vs Red u ctive Elim in a -
tion . The selectivity of the amido intermediate shown
in Scheme 4 for the unusual C-N bond-forming reduc-
tive elimination reaction rather than common â-hydro-
gen elimination is striking. In the chemistry of late
metal amido complexes, the kinetic selectivity of the tri-
o-tolylphosphine palladium system for reductive elimi-
nation over â-hydrogen elimination is unparalleled.
Imine hydrogenations56 involve imine insertion into a
hydride rather than â-hydrogen elimination, but the
preference for insertion is likely to be dictated by
thermodynamic, rather than kinetic, effects. In any
case, the factors that control relative rates for â-hydro-
gen vs reductive elimination are clearly important for
catalysis with amido compounds. Some of the factors
that control the relative rates for these two processes
have been recently reported.55
Con clu sion s. In addition to defining the principal
reactions comprising palladium-catalyzed aromatic ami-
nations, this mechanistic study clearly illustrates that
catalytic organometallic chemistry of the platinum-
group metals involving complexes with metal-carbon
and metal-hydrogen bonds can be extended to similar
catalysis involving platinum-group amides. Although
transition metal alkylamido complexes of low valent
metal centers are rare, they can exist as reactive
intermediates which produce synthetically useful trans-
formations when bound to the proper metal and sur-
rounded by the proper ligand set.3
It is often the case that the complexes lying directly
on a catalytic pathway cannot be isolated or observed.57
In the case of this study, all of the species lying on the
catalytic cycle not only are too reactive to isolate but
are too reactive to even observe by NMR spectroscopic
techniques. The Pd(0) complex on the reaction pathway
is highly unsaturated and contains only one phosphine
ligand. The aryl halide complex in the catalytic cycle
is not the dimeric or amine-ligated complexes that have
been isolated, but is likely to be a monomeric, three-
coordinate aryl halide species. Further, the unusual
amido complex is most likely a neutral, three-coordinate
dialkylamido species that lies uphill from the combina-
tion of aryl halide complex and tin amide and whose
barrier for reaction is lower than that for simple loss of
KBr or LiBr from the observed anionic amido complex.
Thus, the general difficulty in isolating neutral alkyl
or dialkylamido complexes with â-hydrogens to date
should not imply that these complexes cannot act as
important reactive intermediates in catalytic chemistry.
Instead, this high reactivity may make them more prone
to participate in catalysis than late metal amido com-
plexes that have been susceptible to characterization
by conventional spectroscopic techniques.
Exp er im en ta l Section
Gen er a l Da ta . Unless otherwise specified, all reagents
were purchased from commercial suppliers and were used
(49) Scott, W. J .; Stille, J . K. J . Am. Chem. Soc. 1986, 108, 3033-
3040.
(50) Scott, W. J .; Crisp, G. T.; Stille, J . K. J . Am. Chem. Soc. 1984,
106, 4630.
(51) Farina, V.; Kapadia, S.; Krishnan, B.; Wang, C.; Liebeskind,
L. S. J . Org. Chem. 1994, 59, 5905.
(52) Scott, W. J .; Stille, J . K. J . Am. Chem. Soc. 1986, 108, 3033.
(53) Milstein, D.; Stille, J . K. J . Am. Chem. Soc. 1979, 101, 4992.
(54) Amatore, C.; Pfluger, F. Organometallics 1990, 9, 2276-2282.
(55) Hartwig, J . F.; Richards, S.; Baran˜ano, D.; Paul, F. J . Am.
Chem. Soc. 1996, 118, 3626.
(56) Imine hydrogenation catalyzed by rhodium and iridium com-
plexes would be expected to occur through an amide complex. However,
mechanistic data is scarce. A direct observation of imine insertion that
produces an amide complex potentially similar to those in catalytic
systems is the following: Fryzuk, M. D.; Piers, W. E. Organometallics
1990, 9, 986-98.
(57) Halpern, J . Inorg. Chim. Acta 1981, 50, 11.

2804 Organometallics, Vol. 15, No. 12, 1996
Louie et al.
without further purification. Pentane, benzene, toluene, ether,
and THF were distilled under nitrogen over sodium and
benzophenone. Deuterated solvents for use in NMR experi-
ments were dried as their protiated analogs, but were vacuum
transferred from their drying agents. LiCl and LiBr were
heated under vacuum for 1 h prior to use. KF was purchased
in an anhydrous form and was stored and transferred under
nitrogen. Anhydrous NMP solvent was purchased from Ald-
rich. PhNO2 and CH3CN were distilled from CaH2. {Pd[P(o-
Tol)3](Ar)(Br)}2,23,33 {Pd[P(o-Tol)3](Ar)(Br)(NHR2)},33 P(o-Tol)3,58
volume was made up to 3.0 mL with toluene-d8. Samples for
each rate measurement were prepared by weighing the ap-
propriate amount of 3a into a vial and adding 0.54 mL of the
3.0 mL stock solution. The resulting homogeneous solution
was transferred to a screw top NMR tube equipped with a
Teflon septum. Reaction rates were measured by 1H NMR
spectroscopy at 85 °C. Sample tubes were shimmed at room
temperature and removed, and the probe was warmed to 85
°C. The sample was placed into the probe, quickly reshimmed,
and an automated program was initiated that collected single-
pulse experiments with at least 1 min between data collections.
The Bu3SnNMe2 methyl resonance at 2.804 ppm was inte-
grated over the course of at least 3 half-lives. The following
pseudo-first-order rate constants were obtained at different
concentrations of 3a (k, [3a ]): 2.7 × 10-4 s-1, 0.0059 M; 3.3 ×
10-4 s-1, 0.012 M; 9.8 × 10-4 s-1, 0.018 M; 9.4 × 10-4 s-1, 0.024
M; 1.1 × 10-3 s-1, 0.029 M.
59
Bu3SnNMe2,59 and Bu3SnNEt2 were synthesized following
reported procedures.
Unless otherwise noted, all manipulations were carried out
in an inert atmosphere glovebox or by using standard Schlenk
or vacuum line techniques. 1H NMR spectra were obtained
on a GE QE 300 MHz or an Ω 300 Fourier transform
spectrometer. 31P NMR spectra were obtained on the Ω 300
at 121.6 MHz. 1H NMR chemical shifts are reported in units
of parts per million relative to tetramethylsilane and were
referenced to residual protiated solvent. 31P{1H} chemical
shifts are reported in units of parts per million relative to 85%
H3PO4. Infrared spectra were recorded on a MIDAC Fourier
transform spectrometer. Samples for elemental analysis were
submitted to Atlantic Microlab, Inc. GC analyses were
performed on a Hewlett-Packard Series 800 gas chromato-
graph, equipped with a 60 m methylsilicone capillary column.
GC/MS analyses were also conducted on a HP series 800 Gas
Chromatograph, equipped with a 10 m SE 30 column. Quan-
titative analyses were obtained integrating peak areas versus
1,3,5-tri-tert-butylbenzene or naphthalene internal standard.
Response factors relative to the internal standard were
determined by coinjection of known quantities of authentic
product samples with the internal standards.
Or d er in HNMe2. Five samples were prepared by placing
into an NMR sample tube 0.50 mL of a stock solution made
from 24.0 mg (0.0383 mmol) of 3a , 105.9 mg (0.348 mmol) of
P(o-Tol)3, 80.0 µL (0.650 mmol) of p-CH3C6H4Br, 53.0 µL (0.192
mmol) of Bu3SnBr, and 35.0 µL (0.116 mmol) of Bu3SnNMe2
dissolved in 3.0 mL of toluene-d8. Varying amounts of HNMe2
were injected into each sample with a gas tight syringe.
Reactions were monitored at 95 °C. The following pseudo-
first-order rate constants were obtained at different concentra-
tions of HNMe2 (k, [HNMe2]): 7.5 × 10-4 s-1, 0.051 M; 5.0 ×
10-4 s-1, 0.086 M; 4.3 × 10-4 s-1, 0.11 M; 3.0 × 10-4 s-1, 0.15
M; 1.6 × 10-4 s-1, 0.27 M.
Or d er in P (o-Tol)3 a n d p-CH3C6H4Br . Three samples
were prepared by placing into an NMR sample tube 0.50 mL
of a stock solution made from 24.4 mg (0.0389 mmol) of 3a ,
96.3 mg (0.316 mmol) of P(o-Tol)3, 107.6 mg (0.629 mmol) of
p-CH3C6H4Br, 53.0 µL (0.192 mmol) of Bu3SnBr, and 35.0 µL
(0.116 mmol) of Bu3SnNMe2 dissolved in 3.0 mL of toluene-
d8. To one sample was added an additional 67.9 mg (0.223
mmol) of P(o-Tol)3. To another sample was added an ad-
ditional 68.1 mg (0.398 mmol) of p-CH3C6H4Br. Reactions
were monitored at 80 °C. The following pseudo-first-order rate
constants were obtained: (k, [P(o-Tol)3], [p-CH3C6H4Br]): 2.0
× 10-4 s-1, 0.11 M, 0.21 M; 2.1 × 10-4 s-1, 0.57 M, 0.21 M; 2.2
× 10-4 s-1, 0.11 M, 1.0 M.
{P d [P (o-Tol)3](HNMe2)(Br )(p-CH3C6H4)} (3a ). A sus-
pension of 112.4 mg (0.193 mmol) of 2a in toluene (10 mL)
was degasssed, and an excess (ca. 50 equiv) of HNMe2 was
added to the suspension. In minutes, the sparingly soluble
yellow aryl halide complex dissolved to give a pale yellow
solution of 3a . After being stirred for 30 min, the solution was
filtered through Celite to remove any traces of precipitate, and
the resulting solution was concentrated. A pale yellow pre-
cipitate formed upon concentration and this pure 3a (108.0
mg, 89.2%) was isolated by filtration. Alternatively, crystalline
samples of 3a were obtained by slow diffusion of pentane into
a concentrated toluene solution at room temperature. IR
(KBr): 3282 (m), 3244 (w), 3052 (m), 2995 (m), 2963 (m), 2920
(s), 2865 (w), 1589 (s), 1562 (m), 1443 (vs), 1384 (s), 1281 (s),
1199 (m), 1165 (w), 1130 (s), 1070 (m), 1053 (m), 1008 (vs),
895 (m), 795 (vs), 748 (vs), 714 (s), 674 (s), 560 (s), 535 (s), 465
Va r ia tion of Ar yl Br om id e Con cen tr a tion s in th e
Rea ction of p-t-Bu C6H4Br w ith Bu 3Sn NMe2 Ca ta lyzed by
1. Three samples were prepared by placing into an NMR
sample tube 0.70 mL of a stock solution made from 13.5 mg
(0.0189 mmol) of 1, 49.3 µL (0.280 mmol) of p-t-BuC6H4Br, and
57.5 mg (0.189 mmol) of P(o-Tol)3 dissolved in 2.1 mL of
benzene-d6. To one sample was added an additional 16.4 µL
(0.0940 mmol) of p-t-BuC6H4Br. To another sample was added
an additional 32.9 µL (0.190 mmol) p-t-BuC6H4Br. Bu3-
SnNMe2 (10.5 µL, 0.0349 mmol) were injected into the samples
immediately before data collection at 60 °C. Reactions were
monitored to 90% completion. Although first-order plots were
slightly curved due to the absence of added tin bromide,
inspection of concentrations of the tin amide at different time
points during these reactions showed essentially identical
conversions at each time point.
Va r ia tion of P (o-Tol)3 Con cen tr a tion s in th e Rea ction
of p-t-Bu C6H4Br w ith Bu 3Sn NMe2 Ca ta lyzed by 1. Three
0.70 mL samples were prepared that each contained 4.5 mg
(0.0063 mmol) of 1, 21.0 µL (0.120 mmol) p-t-BuC6H4Br, and
60.1 µL (0.200 mmol) of Bu3SnNMe2 in toluene-d8. Then 11.0
mg (0.0361 mmol), 22.0 mg (0.0723 mmol), and 44.0 mg (0.145
mmol) of P(o-Tol)3 were added to individual samples. Reac-
tions were monitored at 75 °C.
(s) cm-1
.
1H NMR (C7D8, 80 °C): δ 7.94 (broad s, 2H), 6.80-
7.17 (m, 12H), 6.51 (d, J ) 7.8 Hz, 2H), 3.45 (broad s, 1H),
2.29 (broad s, 9H), 2.20 (broad s, 6H), 2.02 (s, 3H). 31P{1H}
NMR (C7H8, 80 °C): δ 28.6 (s). 31P{1H} NMR (C7H8, 20 °C):
δ 28.7 (s). Anal. Calcd for C30H35NPdPBr: C, 57.48; H, 5.63;
N, 2.23. Found: C, 57.58; H, 5.70; N, 2.31.
Kin etic Stu d ies. Most of the samples for kinetic experi-
ments were prepared in a similar fashion to each other.
Deviations from normal setup procedures are noted. The
procedure used for obtaining the reaction order in palladium
catalyst concentration in the reaction of p-CH3C6H4Br with
Bu3SnNMe2 catalyzed by 3a will be used as the example.
Concentrations, reaction temperatures, and rate constants for
subsequent runs are provided after the detailed description
of this procedure.
Kin et ic Mea su r em en t s of R ea ct ion b et w een p --
CH3C6H4Br a n d Bu 3Sn NMe2 Ca ta lyzed by Am in e Com -
p lex 3a . Or d er in P a lla d iu m Ca ta lyst. Into a 3.0 mL
volumetric flask were added 82.3 mg (0.270 mmol) of P(o-Tol)3,
98.2 µL (0.355 mmol) of Bu3SnBr, 33.0 µL (0.110 mmol) of Bu3-
SnNMe2, and 121.2 mg (0.709 mmol) of p-CH3C6H4Br and the
R ea ct ion of 1 w it h p-Bu C6H 4Br in t h e P r esen ce of
Bu 3Sn Br . First, 5.0 mg (0.0070 mmol) of 1, 11.9 mg (0.0391
mmol) of P(o-Tol)3, and 2.6 mg (0.016 mmol) of 1,3,5-tri-
methoxybenzene were dissolved in 0.5 mL of benzene-d6. A
1H NMR spectrum was obtained and the quantity of 1 vs 1,3,5-
trimethoxybenzene internal standard was determined by
integration. A mixture of 20.2 µL (0.121 mmol) of p-BuC6H4-
(58) Ziegler, C. B., J r.; Heck, R. F. J . Org. Chem. 1978, 43, 2941-
2946.
(59) J ones, K.; Lappert, M. F. J . Chem. Soc. 1965, 1944-1951.

Catalysis with Pt-Group Alkylamido Complexes
Organometallics, Vol. 15, No. 12, 1996 2805
Br and 8.0 µL (0.029 mmol) of Bu3SnBr was added. The
reaction was heated at 60 °C, and integration of a second 1H
NMR spectrum obtained at this time showed that formation
of 2c had occurred in 88% yield.
placed into an airtight syringe. The NMR sample tube was
placed into an NMR probe at -70 °C before addition of
silylamide and 31P and 1H NMR spectra were obtained. The
sample was removed and placed into a -78 °C bath. The
silylamide base was then added by syringe, and the sample
was shaken before replacing it in the NMR probe. 1H and 31P
NMR spectra were recorded at -70 °C and the sample was
monitored by increasing the temperature by 10 °C intervals
until formation of arylamine and L2Pd0 had occurred.
Room Tem p er a tu r e Ad d ition s. Rea ction of 2c w ith
Lith iu m Am id es. Into a screw top vial equipped with a
Teflon lined septum was added 11.2 mg (0.0183 mmol) of 2c.
This complex was dissolved in 0.5 mL of THF in the presence
of 24.3 mg (0.0798 mmol) of P(o-Tol)3 by gentle warming. Then
50.0 µL of a solution or slurry of LiNRR′ that would be 0.70
M if fully dissolved (NMe2, NEt2, NH-t-Bu, NHPh, or NPh2)
was injected after the solution cooled to room temperature.
Reaction mixtures turned from yellow to red. After 20 min at
ambient temperatures, 10 µL of solution containing an internal
standard (0.016 M naphthalene or 1,3,5-tri-tert-butylbenzene
in toluene) was added to the sample. Reactions were analyzed
for arylamine formation by GC and GC/MS.
Sa lt Effects. Catalytic reactions with various salt additives
were prepared in a similar manner to the procedure described
for competition reactions. Each reaction contained 13.5 mg
(0.0444 mmol) of P(o-Tol)3, 22.3 mg (0.0910 mmol) of 1,3,5-
tri-tert-butylbenzene, 85.0 µL (0.253 mmol) of Bu3SnNEt2, and
35.0 µL (0.198 mmol) of p-BuC6H4Br in 2 mL of toluene or
THF solvent. All possible pairs of the following Pd catalysts
and salts were combined with the above mixtures and reacted
at 75 °C for 12 h: 6.2 mg (0.0087 mmol) of 1; 7.6 mg (0.0087
mmol) of Pd[P(o-Tol)3]2Br2; 11.5 mg (0.198 mmol) of KF; 8.0
mg (0.20 mmol) of LiCl; 7.6 mg (0.088 mmol) of LiBr.
Reactions were monitored by GC.
Cop p er Ad d itives. A stock solution containing 3.9 mg
(0.0054 mmol) of 1, 9.1 mg (0.030 mmol) of P(o-Tol)3, 75.0 µL
(0.224 mmol) of Bu3SnNEt2, 20.0 µL (0.113 mmol) of p-BuC6H4-
Br, and 13.9 mg (0.108 mmol) of naphthalene was prepared
in 2 mL of THF. A 0.50 mL amount of the stock solution was
added to the following quantities of individual copper re-
agents: 3.7 mg (0.037 mmol) of CuCl, 4.4 mg (0.031 mmol) of
CuBr, and 6.3 mg (0.033 mmol) of CuI. Reactions were heated
to 75 °C for 4 h and monitored by GC. The remaining 0.5 mL
stock solution was heated simultaneously with the copper
reactions as a control experiment and used to compare solvent
effects (vide supra).
Rea ction Ra te in th e P r esen ce of N,N′-Dim eth yla -
n ilin e P r od u ct. A sample containing 4.4 mg (0.0062 mmol)
of 1, 19.5 mg (0.0637 mmol) of P(o-Tol)3, 49.2 µL (0.280 mmol)
of p-t-BuC6H4Br, and 8.0 µL of freshly distilled N,N′-dimethy-
laniline in 0.70 mL benzene-d6 was prepared. Reaction with
10.2 µL (0.0339 mmol) of Bu3SnNMe2 at 60 °C gave an
observed rate constant of 4.8 × 10-4 s-1
.
Rea ction Ra te in P ola r Solven ts. A sample containing
0.98 mg (0.0014 mmol) of 1, 2.3 mg (0.0075 mmol) of P(o-Tol)3,
18.8 µL (0.0537 mmol) of Bu3SnNEt2, 5.0 µL (0.028 mmol) of
p-BuC6H4Br, and 3.5 mg (0.027 mmol) of naphthalene was
prepared in 0.5 mL of THF. A similar sample containing 1.0
mg (0.0014 mmol) of 1, 2.1 mg (0.0070 mmol) of P(o-Tol)3, 18.8
µL (0.0559 mmol) of Bu3SnNEt2, 5.0 µL (0.028 mmol) of
p-BuC6H4Br, and 4.4 mg (0.028 mmol) of naphthalene was
prepared in 0.5 mL of toluene. These samples were also used
as controls for the effect of copper additives (vide infra).
Conversion of aryl halide to arylamine was monitored by GC
at 75 °C at different reaction times over the course of 2-5 h.
A sample containing 3.1 mg (0.0043 mmol) of 1, 20.0 mg
(0.0657 mmol) of P(o-Tol)3, and 3.9 µL (0.022 mmol) of p-t-
BuC6H4Br in 0.50 mL of THF-d8 was prepared. Reaction with
37.5 µL (0.125 mmol) of Bu3SnNMe2 at 60 °C gave an observed
rate constant of 3.0 × 10-4 s-1. Reactions were similarly set
up in PhNO2, NMP, and CH3CN solvents although no rate
constants were obtained.
Rela tive Ra te Stu d y w ith Su bstitu ted Ar yl Br om id es.
Samples were prepared containing 4.5 mg (0.0063 mmol) of
1, 0.284 mmol of the ArBr, and 19.5 mg (0.0641 mmol) of P(o-
Tol)3 in 0.70 mL of benzene-d6. Bu3SnNMe2 (5.3 µL, 0.018
mmol) was injected into each tube prior to data collection. The
following observed rate constants (k) were obtained at 60 °C
for various substituted aryl bromides: ArBr, 2.4 × 10-4 s-1
p-Me2NC6H4Br; 9.0 × 10-4 s-1, p-CF3C6H4Br; 2.9 × 10-4 s-1
p-MeOC6H4Br; 6.5 × 10-4 s-1, p-FC6H4Br.
,
,
Qu a n t it a t ive R ela t ive R a t es for R ea ct ion of
Bu 3Sn NMe2 a n d Bu 3Sn NEt2. To 0.50 mL of a 1.2 mL stock
solution containing 9.2 mg (0.013 mmol) of 1, 16.5 mg (0.0542
mmol) of P(o-Tol)3, 35.2 mg (0.206 mmol) of p-CH3C6H4Br, and
15.0 µL (0.0543 mmol) of Bu3SnBr in toluene-d8 was added
either 12.0 µL (0.0399 mmol) Bu3SnNMe2 or 13.0 µL (0.0388
mmol) Bu3SnNEt2. In one case, the disappearance of Bu3-
SnNMe2 was monitored and in the other, the appearance of
p-CH3C6H4NEt2 was monitored, both at 75 °C over 3 half-lives.
Pseudo-first-order rate constants of 6.6 × 10-4 s-1 and 4.9 ×
10-4 s-1 were obtained for the reactions with Bu3SnNMe2 and
Bu3SnNEt2, respectively, giving a relative ratio of 1.4.
Com p et it ion St u d y b et w een Bu 3Sn NMe2 a n d
Bu 3Sn NEt2. Into a screw top vial equipped with septum were
added 1.2 mg (0.0017 mmol) of 1, 2.3 mg (0.0076 mmol) of P(o-
Tol)3, 5.0 µL (0.028 mmol) of p-BuC6H4Br and between 1.0 and
3.0 mL of toluene. Bu3SnNMe2 (70.0 µL, 0.233 mmol) and Bu3-
SnNEt2 (80.0 µL, 0.239 mmol) were syringed into the medium.
The vial was sealed and warmed to 70 °C for 2 h. The ratio of
arylamine products was analyzed by GC. The average of two
runs was found to be 2.9 ( 0.4.
Another stock solution containing 4.1 mg (0.0057 mmol) of
1, 8.5 mg (0.028 mmol) of P(o-Tol)3, 75.0 µL (0.224 mmol) of
Bu3SnNEt2, 20.0 µL (0.131 mmol) of p-BuC6H4Br, and 14.1 mg
(0.110 mmol) of naphthalene was prepared in 2 mL of toluene.
A 0.5 mL amount of the stock solution was added to 4.0 mg
(0.040 mmol) of CuCl, 4.4 mg (0.031 mmol) of CuBr, or 6.3 mg
(0.033 mmol) of CuI. Reactions were heated to 75 °C for 4 h
and monitored by GC. The remaining 0.5 mL stock solution
was heated simultaneously as a control experiment and used
to compare solvent effects (vide supra).
Ack n ow led gm en t. We are grateful to DuPont for
a Young Professor Award, Union Carbide for an In-
novative Recognition Award, the Dreyfus Foundation
for a New Faculty Award, the National Science Founda-
tion for a Young Investigator Award, and to the Exxon
Education Foundation, Rohm and Haas, and Bayer
Pharmaceuticals for generous support. We also grate-
fully acknowledge a loan of palladium dichloride from
J ohnson Matthey Alfa/Aesar.
In d ep en d en t Syn th esis. Amine complex 3b {[P(o-Tol)3]-
(NHEt2)Pd[p-BuC6H4](Br)} was used for synthetic studies at
low temperatures because its 31P NMR spectrum displays one
dominant resonance in THF solvent.
Low Tem p er a tu r e Ad d ition s. Complex 3b (6.0 mg,
0.0086 mmol) was dissolved in 0.5 mL of THF or THF-d8 and
placed into a screw-topped NMR sample tube equipped with
a Teflon lined septum. Then 2 equiv of either LiN(SiMe3)2 or
KN(SiMe3)2 was dissolved into 0.2 mL of THF or THF-d8 and
OM960188O