Palladium-catalyzed amination of aryl bromides and aryl triflates using diphosphane ligands: A kinetic study

Article abstract of DOI:10.1002/1521-3765(20010119)7:2<475::AID-CHEM475>3.0.CO;2-6

[Pd(P-O-P)(Ar)]⁺ complexes with ligands that have wide bite angles are active catalysts for the coupling of aniline derivatives with aryl triflates. Kinetic studies show that for these systems a fast equilibrium that involves coordination of the amine precedes the deprotonation, which is the rate-limiting step of the reaction. This reaction is faster for compounds with a smaller P-Pd-P angle. When halide salts are present, the base sodium tert-butoxide is activated and adds to the palladium complex. This rate-limiting step is preceded by a fast equilibrium that involves decoordination of the halide. The initial reaction rate is faster for compounds with a larger P-Pd-P angle. This is due to the closer proximity of the oxygen to the Pd center, and this assists in the dissociation of the halide.

Full text of DOI:10.1002/1521-3765(20010119)7:2<475::AID-CHEM475>3.0.CO;2-6

FULL PAPER
Palladium-Catalyzed Amination of Aryl Bromides and Aryl Triflates
Using Diphosphane Ligands: A Kinetic Study
Yannick Guari, Gino P. F. van Strijdonck, Maarten D. K. Boele, Joost N. H. Reek,
Paul C. J. Kamer, and Piet W. N. M. van Leeuwen*[a]
‡
ꢀ ꢀ
Abstract: [Pd(P O P)(Ar)] complexes with ligands that have wide bite angles are
active catalysts for the coupling of aniline derivatives with aryl triflates. Kinetic
studies show that for these systems a fast equilibrium that involves coordination of
the amine precedes the deprotonation, which is the rate-limiting step of the reaction.
Keywords: amination
neous catalysis ´ kinetics ´ palladi-
um ´ phosphanes
´
homoge-
ꢀ
ꢀ
This reaction is faster for compounds with a smaller P Pd P angle. When halide salts
are present, the base sodium tert-butoxide is activated and adds to the palladium
complex. This rate-limiting step is preceded by a fast equilibrium that involves
decoordination of the halide. The initial reaction rate is faster for compounds with a
ꢀ
ꢀ
larger P Pd P angle. This is due to the closer proximity of the oxygen to the Pd
center, and this assists in the dissociation of the halide.
complex [Pd(PP)(Ar)(X)] (1) (step I). This complex can react
with sodium tert-butoxide to form complex 3. The formation
of [Pd(PP)(OR)(Ar)] (3) could proceed either by an associa-
tive or a dissociative (steps II and III) pathway. Since a
hemilabile terdentate ligand stabilizes the cationic intermedi-
ate 2, the dissociative mechanism could be preferred with the
use of a potentially terdentate ligand. Reaction of 3 with the
amine (step IV) will give the amido complex 5. Reductive
elimination of the product (step V) will regenerate the active
intermediate 6.
Introduction
During the past five years, extensive research devoted to the
development of palladium-catalyzed carbon-nitrogen bond
formation has led to efficient systems that offer considerable
advantages over the classical methods (i.e. nonactivated
substrates can be used, and neither highly polar solvents nor
severe reaction conditions are required).[1] In the search for
ligands that provide more active catalysts, improvements have
recently been made by using bidentate phosphane,[2] amino-
phosphane,[3] and bulky electron-rich phosphane[4] ligands.
The elegant work of both Hartwig and Buchwald has
elucidated part of the mechanism of the amination reactions.
These studies focused on monophosphane Pd species and cis
chelating diphosphane Pd species by using catalysts prepared
in situ.
‡
Alternatively, complex [Pd(PP)(Ar)] (2) could react with
the amine (step III') to afford complex 4. Deprotonation by
the base (step IV') will afford the amido complex 5, which
again will undergo reductive elimination. If the neutral
complex 1 reacts with the amine, a similar pathway can be
obtained that involves a neutral pentacoordinated Pd(amine)
species and an anionic pentacoordinated amido complex.
Oxidative addition of aryl triflates will directly lead to the
formation of cationic complex 2.
Possible catalytic pathways for the amination of aryl halides
are depicted in Scheme 1. The catalytic cycle starts with the
oxidative addition of an aryl halide to [Pd(P P)] to give
0
ꢀ
Hartwig proposed a catalytic cycle for cis chelating
diphosphane Pd catalysts in the amination of aryl halides.[5]
In this catalytic cycle, oxidative addition of an aryl halide to
[a] Prof. Dr. P. W. N. M. van Leeuwen, Dr. Y. Guari
Dr. G. P. F. van Strijdonck, Dr. M. D. K. Boele
Dr. J. N. H. Reek, Dr. P. C. J. Kamer
0
ꢀ
[Pd(P P)] (step I) gives complex 1, [Pd(PP)(Ar)(X)]. Suc-
Institute of Molecular Chemistry
cessive reactions with base (step II) and amine (step III) lead
Department of Inorganic Chemistry, Universiteit van Amsterdam
Nieuwe Achtergracht 166, 1018WV Amsterdam (The Netherlands)
Fax : (‡ 31)20-5256456
to amido complex 5, after which reductive elimination of
0
ꢀ
ArNRR' regenerates [Pd(P P)] (6). An alternative route,
E-mail: pwnm@anorg.chem.uva.nl
suggested by Buchwald, is based on the reaction of the amine
with [Pd(PP)(Ar)(X)] (1) to afford a pentacoordinate com-
plex followed by deprotonation by the base to form the amido
complex.[6]
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author. Plots that
show the zeroth-order dependency of the remaining reagents are
available as Supporting information.
Chem. Eur. J. 2001, 7, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0475 $ 17.50+.50/0
475

P. W. N. M. van Leeuwen et al.
FULL PAPER
atom is coordinated to the met-
al center. Isolation of cis com-
plexes 2'a, b, and g required
acetonitrile.
ArNRR'
ArX
ArNRR'
Pd0(P-P)
Pd0(P-P)
6
6
I
I
V
V
In order to interpret the li-
gand effects and to determine
the origin of the difference in
performance found for DPE-
phos and Xantphos ligands in
the palladium-catalyzed amina-
tion reaction, we used com-
plexes 1 and 2 in an amination
reaction. A kinetic study was
performed to gain further in-
sight into the details of the
catalytic cycle.
Ar
X
P
P
Pd
1
Ar
Ar
P
P
P
P
II
II
Pd
5
Pd
5
Ar
P
P
Pd+
2
NRR'
NRR'
III
III'
NaOtBu
NHRR'
IV
IV'
HOtBu
HOtBu
Ar
Ar
P
P
P
P
Pd
Pd+
OtBu
NHRR'
-OtBu
3
4
HNRR'
Scheme 1. Possible pathways for the palladium-catalyzed carbon-nitrogen bond formation.
We previously described the use of the ligand Xantphos
(9,9-dimethyl-4,6-bis(diphenylphosphano)xanthene, e) in the
palladium-catalyzed amination of aryl bromides.[7] With
Pd(OAc)2 as the Pd source, this system establishes the
efficient and selective coupling of aryl bromides with primary
aliphatic amines, anilines, and cyclic secondary amines. The
use of DPEphos (bis[2-(diphenylphosphano)phenyl]ether, a)
as a ligand was reported to be less general and less efficient
than other Xantphos-type ligands (see Scheme 2).
Results
As a model reaction we used the coupling of 2-methoxyaniline
(o-anisidine) with bromobenzene and phenyl triflate. Sodium
tert-butoxide was used as a base and toluene as a solvent.
Since it is known that dba (dibenzylidene acetone) can
interfere with the oxidative addition[10] and we wanted to
exclude any rate effect of the formation of a catalytic species
from a precatalyst (e.g. reduction of PdII to Pd0), we used
compounds 1 and 2, as proposed intermediates of the catalytic
cycle, rather than using in situ prepared catalysts from
[Pd2(dba)3] or Pd(OAc)2 and ligand.
R
X
Y
R
ligand = a-f
ligand = g-j
PPh2
Fe
PPh2
Catalysis with neutral complexes (1): In Table 1, the results for
the arylation of 2-methoxyaniline catalyzed by compounds 1
at room temperature are shown. In all reactions, only the
monoarylated product was observed. Under these reaction
conditions, the PdII compounds with ligating monophos-
phanes (entry 8 and 9) show no conversion. Also for
[Pd(dppe)(Ar)(Br)] (1,2-bis(diphenylphosphano)ethane) (1h),
no activity was observed (entry 7). The compounds with the
ligands enforcing a cis coordination mode gave low conver-
PPh2
PPh2
Ph2P
PPh2
Scheme 2. Structures of phosphane ligands: a ˆ DPEphos, X ˆ H,H, Yˆ
O, and R ˆ H; b ˆ Homoxantphos, X ˆ CH2CH2, Yˆ O, and R ˆ H; c ˆ
Sixantphos, X ˆ Si(CH3)2, Yˆ O, and R ˆ H; d ˆ Thixantphos, X ˆ S, Yˆ
O, and R ˆ CH3; e ˆ Xantphos, X ˆ C(CH3)2, Yˆ O, and R ˆ H; f ˆ
Thioxantphos, X ˆ C(CH3)2, Yˆ S, and R ˆ H; g ˆ dppf; h ˆ dppe; i ˆ
PPh3; j ˆ P(o-tolyl)3.
OTf
Br
Pd
P
P
Recently, we have prepared a series of [palladium(ii)-
(p-C6H4CN)] complexes with different diphosphane ligands
(see Scheme 3). They have been fully characterized and
details will be presented elsewhere.[8]
PP = c-f
P
Pd
P
CN
CN
Y
PP = a-e, g
1
2
The structures of complexes 1a, b, and g show a chelating
phosphane that accommodates a common cis coordination
mode.[9] Complexes 1c ± e, however, show a trans coordina-
tion mode. We assume that the oxygen is forced into proximity
of the Pd center by the trans coordination mode of the
CH3
C
+
N
PP = a, b, g
P
Pd
P
CN
ꢀ
phosphane groups, and this leads to a weak Pd O interaction.
In solution, compounds 1c ± e show a cis/trans isomerization.
2'
‡
Furthermore, cationic complexes [Pd(PP)(p-C6H4CN)] 2c ± f
Scheme 3. [PdII(PP)(p-C6H4CN)] complexes (PP): 1a ˆ DPEphos; 1b ˆ
Homoxantphos; 1c ˆ Sixantphos; 1d ˆ Thixantphos; 1e ˆ Xantphos;
1g ˆ dppf; 2c ˆ Sixantphos, and Yˆ O; 2d ˆ Thixantphos, and Yˆ O;
2e ˆ Xantphos, and Yˆ O, 2 f ˆ Thioxantphos, and Yˆ S; 2'a ˆ DPEphos;
2'b ˆ Homoxantphos; 2'g ˆ dppf.
have been synthesized and fully characterized. These com-
plexes adopt an almost square planar geometry around the
metal center with the ligand acting as a ªpincerº, and the
phosphane groups are in trans orientation, and the oxygen
476
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0476 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 2

Palladium-Catalyzed Amination of Aryl Bromides and Aryl Triflates
475±482
Table 1. Catalytic amination of bromobenzene at 258C with complexes 1.[a]
Effect of halide salts: Since during the reaction of bromo-
benzene with 2-methoxyaniline sodium bromide is formed, we
tested the effect of the halide salt concentration on the activity
of compound 2e in the amination of phenyl triflate. Table 3
shows that the activity of the catalyst in the model reaction is
largely dependent on the concentration of halides present. In
general, the initial activity was higher when halide salt was
[b]
complex
TOFini
conversion[c]
[%]
natural bite angle[d]
[molmolꢀ1 hꢀ1
]
1
2
3
4
5
6
7
8
9
1a
1b
1c
1d
1e
1g
1h
1i[e]
1j[f]
22
27
80
86
140
0
0
0
0
3
11
40
47
96
1
0
0
0
102
102
109
110
111
96
85
±
±
Table 3. The effect of added halide salts in the amination of phenyl triflate
at 258C with complex 2e.[a]
[additive] [mm]
TOFini[b] [molmolꢀ1 hꢀ1
]
yield[c] [%]
[a] Conditions: see Experimental Section. [b] TOFini is the amount of
product formed per molPd per hour after six minutes. [c] Based on
bromobenzene (o-anisidine and product yield give similar results) after 1 h.
[d] Values taken from ref. [11]. For a recent review about bite angles, see
ref. [11]. [e] [Pd(PPh3)2(p-C6H4CN)(Br)] was used. [f] Dimeric compound
[Pd(P(o-tolyl)3)(p-C6H4CN)(m-Br)]2 was used.
1
2
3
4
5
6
7
8
9
no
160
290
257
260
384
355
640
624
169
224
238
208
35
59
42
76
79
67
68
49
72
75
77
74
3.9 (NaCl)
27.4 (NaCl)
165.6 (NaCl)
0.5 (Bu4NBr)
2.0 (Bu4NBr)
29.4 (Bu4NBr)
200 (Bu4NBr)
0.8 (Bu4NI)
2.5 (Bu4NI)
47 (Bu4NI)
276 (Bu4NI)
sions (entries 1, 2, and 6), whereas the complexes with wide
ꢀ
ꢀ
P Pd P angles showed excellent conversions after one hour
under these very mild conditions. The observed reaction rates
for these ligands follow the order c < d < e.
10
11
12
[a] Conditions: see Experimental Section. [b] TOFini is the amount of
product formed per molPd per hour after six minutes. [c] Based on phenyl
triflate after 1 h.
Catalysis with cationic complexes (2): In Table 2, the results of
the reaction between 2-methoxyaniline and phenyl triflate
catalyzed by the cationic complexes 2 are presented. The
trend observed for the reaction rate with the neutral
complexes is reversed. Compound 2 f, which only exists in
the cationic form due to the strong coordination of the soft
sulfur atom,[8] is not active in the amination reaction under the
conditions tested.
added. At higher salt concentrations, the initial reaction rate
slowed down again. The effect was most prominent for the use
of bromide. The values found for the yield after one hour were
obscured by the drop in base concentration due to the
butanolysis reaction.
The halide salt concentration has a dramatic effect on
the reaction rate and kinetics of the amination of bromo-
benzene. Figure 1 shows a nonlinear first-order plot of the aryl
halide concentration in the reaction with 2-methoxyaniline
catalyzed by 1e. From these results, it is clear that the reaction
rate actually increases during the course of the reaction.
Addition of an excess of tetrabutylammonium bromide
(50 equivalents relative to complex 1e) provided a linear
first-order plot.
Table 2. Amination of phenyl triflate at 258C with complexes 2.[a]
[b]
complex
TOFini [molmolꢀ1 hꢀ1
]
conversion[c] [%]
yield [%]
1
2
3
4
2c
2d
2e
2 f
222
200
160
0
76
66
50
0
50
44
35
0
[a] Conditions: see Experimental Section. [b] TOFini is the amount of
product formed per molPd per hour after six minutes. [c] Based on phenyl
triflate after 1 h.
Kinetic experiments: Since the data for the amination of
phenyl triflate were obscured by the cleavage reaction, we
were not able to analyze the reaction profile. To be able to
determine the rate-limiting step, we performed some single-
point kinetic experiments. The observed initial rate constants,
kini, for reactions with different initial concentrations of PdII
complex, 2-methoxyaniline, bromobenzene or phenyl triflate,
and sodium tert-butoxide were determined. The concentration
of the other reactants was kept constant by using them in large
excess.
The conversion of phenyl triflate is higher than the
conversion of the amine and the product yield. This is due
to reaction of sodium tert-butoxide with phenyl triflate that
forms sodium phenoxide and butyl triflate. This butanolysis
reaction has been observed before.[12] This cleavage of triflates
can be avoided by the slow addition of the aryl triflate or the
use of a non-nucleophilic base. The latter solution, however, is
not generally applicable, since for some aminations, the
reaction only proceeds when sodium tert-butoxide is used.
Since under standard conditions the butanolysis reaction is
slower than the amination reaction (less than 3% after six
minutes), the kinetics of the cationic reaction is only
influenced at high conversions.
Neutral complexes: Kinetic data for the reaction between
2-methoxyaniline and bromobenzene catalyzed by complex-
es 1a and 1e were obtained by monitoring the formation of
product by GC. Since complex 1a is not very active at 258C,
the reactions were performed at 408C.
‡
The cationic complexes [Pd(PP)(Ar)(CH3CN)] (2')
showed no conversion in the amination reaction under the
conditions employed. In general, addition of acetonitrile to
amination reaction mixtures inhibits the formation of product.
The observed initial rate constant is independent of the
concentration of 2-methoxyaniline (0.09 ± 0.74m) and the
Chem. Eur. J. 2001, 7, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0477 $ 17.50+.50/0
477

P. W. N. M. van Leeuwen et al.
FULL PAPER
Figure 3. Plot of the rate[a] versus [Pd]1/2 with [1e] and [1a] for the reaction
of 2-methoxyaniline with bromobenzene in toluene at 313 K.[b] [a] rate ˆ
d[product]/dt, measured after six minutes. [b] For experimental details: see
Experimental Section.
Effect of halide: Kinetic experiments with 1e as the catalyst in
the presence of 50 equivalents of tetrabutylammonium bro-
mide (in relation to palladium) showed a first-order depen-
dency on the palladium concentration (0 ± 4.15mm, Tˆ 298 K,
see Figure 4).
Figure 1. Plot of concentration versus time for the reaction of 2-methoxy-
aniline with bromobenzene catalyzed by complex 1e (top) and first-order
plot of the decay of bromobenzene with Brꢀ (50 equiv) and without
additional Bu4NBr (bottom).
concentration of bromobenzene (0.23 ± 0.62m). Figure 2
shows the linear dependence of kini on the sodium tert-
butoxide concentration (0 ± 0.65m). The reaction rate increas-
es approximately linearly with the square root of the
Figure 4. Plot of the rate[a] versus [Pd] for the reaction of 2-methoxyaniline
with bromobenzene in toluene at 298 K catalyzed by 1e in presence of
50 equivalents of tetrabutylammonium bromide.[b] [a] rate ˆ d[product]/dt,
measured after six minutes. [b] For experimental details: see Experimental
Section.
Cationic complexes: With complex 2e, the reaction rate
increases linearly with the palladium concentration (0 ±
4mm, see Figure 5). This reaction shows an approximate
first-order in 2-methoxyaniline and sodium tert-butoxide (see
Figure 6), but the reaction rate is independent of the aryl
triflate concentration.
NMR experiments: In the 31P{1H}NMR spectrum of
[Pd(Xantphos)(p-C6H4CN)(Br)] (1e) in chloroform, a singlet
was observed at d ˆ 9.4, whereas the cationic compound
[Pd(Xantphos)(p-C6H4CN)]OTf (2e) showed a singlet at d ˆ
18.2. A mixture of 1e and 2e resulted in a broad peak between
d ˆ 9.4 and 18.2. At lower temperatures (190 K), these peaks
decoalesced to their initial positions. Addition of an excess of
sodium tert-butoxide to this mixture gave rise to a signal at
Figure 2. Plot of the kini[a] versus [NaOtBu] for the reaction of 2-methoxy-
aniline with bromobenzene in toluene at 313 K catalyzed by complexes 1e
and 1a.[b] [a] Measured after six minutes. [b] For experimental details: see
Experimental Section.
palladium concentration (0 ± 6.77mm, see Figure 3). No for-
mation of palladium black was observed under the conditions
used.
478
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0478 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 2

Palladium-Catalyzed Amination of Aryl Bromides and Aryl Triflates
475±482
Discussion
From the kinetic results, obtained with the cationic com-
plexes 2 and the neutral complexes 1, we can conclude that
the reaction pathway shifts to a different mechanism when
halide salts are present.
Hartwig et al. have reported the synthesis of palladium(ii)-
ꢀ
amido complexes and showed that for the C N bond forming
reductive elimination of monomeric phosphane ligated com-
plexes two competing pathways are operative.[13] Cis chelated
complexes of diphosphanes like dppf (1,1'-bis(diphenylphos-
phano)ferrocene) eliminate the product amine directly from a
four-coordinate [PdII(PP)(amido)(aryl)] complex. The kinetic
data obtained in this study show that the reductive elimination
is not the rate-limiting step, since the reaction rate shows a
first-order dependence in base.
Figure 5. Plot of the rate[a] versus [Pd] for the coupling of 2-methoxyani-
line with phenyltrifluoromethane sulfonate in toluene at 298 K catalyzed
by 2e.[b] [a] rate ˆ d[product]/dt, measured after six minutes. [b] For
experimental details: see Experimental Section.
Recently, Buchwald reported that with Pd(OAc)2/BINAP
(2,2'-bis(diphenylphosphano)-1,1'-binaphthyl) as precatalyst,
the oxidative addition is the rate-limiting step.[6c] From the
presented results (kini is independent of [bromobenzene] or
[phenyl triflate]), however, we can conclude that by using our
system the oxidative addition is not the rate-limiting step
either. A similar result has been observed in a kinetic study
dedicated to the palladium-catalyzed arylation of alkenes.[14]
Catalysis with cationic complexes (2): For the cationic
compounds 2 we found the rate law expressed in Equa-
tion (1).
Figure 6. Plots of kini[a] versus [NaOtBu] and [o-anisidine] for the coupling
of o-anisidine with phenyltrifluoromethane sulfonate in toluene at 298 K
catalyzed by 2e.[b] [a] Measured after six minutes. [b] For experimental
details: see Experimental Section.
v ˆ k[2][NaOtBu][o-NH2C6H4OCH3]
(1)
This equation is consistent with a mechanism in which the
cationic starting compound (2) is in fast equilibrium with
compound 4. In the rate-limiting step, this compound is then
deprotonated by the base to give compound 5 (see Scheme 4).
d ˆ 3.8. This signal was also observed when 1e was reacted
with tert-butoxide.
We studied the reaction by NMR spectroscopy under
catalytic conditions to observe the resting state of the catalyst.
To this end, a substoichiometric amount of sodium tert-
butoxide was used in the reaction of 2-methoxyaniline with
bromobenzene, with 1e as the catalyst. At low temperatures,
two broad signals at d ˆ 9.4 and 19.8 can be observed in
31P{1H}NMR spectroscopy (see Figure 7). When the reaction
takes place (at higher temperatures), these signals are still
observed. Once the base had been consumed and the coupling
reaction stopped, only the signal at d ˆ 9.4 was observed.
+
tBuO-
+
P
O Pd
P
NH2
P Pd
NH
P Pd
P
P
2
4
5
Scheme 4. Rate-limiting part of a proposed mechanism for the amination
of aryl triflates.
Subsequently, the proposed pathway of reductive elimination
and oxidative addition to give compound 2 is followed (see
Scheme 1). The presented kinetic data, however, do not
distinguish this pathway from a mechanism in which a
pentacoordinated species [Pd(PP)(Ar)(amine)(OtBu)] is in-
volved. An alternative explanation for the observed rate law
would be a relatively slow reductive elimination from com-
pound 5, preceded by a fast equilibrium between 2, 4, and 5.
Since reductive elimination is faster with increasing bite angle,
the observed bite angle[11] dependence (see Table 2) is not in
agreement with this latter alternative mechanism.
Figure 7. 31P{1H}NMR spectra of a mixture of complex 1e, bromobenzene,
2-methoxyaniline, and sodium tert-butoxide in [D8]toluene recorded
successively at a) 193 K, b) 273 K, c) 353 K, and d) 233 K.
Going from a smaller bite angle (ligand c) to a larger bite
angle (ligand e), the rate of the reaction decreases. We have
Chem. Eur. J. 2001, 7, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0479 $ 17.50+.50/0
479

P. W. N. M. van Leeuwen et al.
FULL PAPER
shown previously that compounds 1 exist as a mixture of cis
and trans isomers in fast equilibrium.[8] Compounds with a
larger natural bite angle prefer the trans isomer, whereas
ligands with a smaller natural bite angle have a preference for
the cis isomer. Ligands with a larger bite angle stabilize
compounds 2, and thus they shift the equilibrium between 2
and 4 towards the inactive intermediate 2. Compound 2 f, with
a strongly coordinating sulfur atom, shows no activity, since
the amine is not capable of replacing the sulfur. When
acetonitrile is added to the reaction mixture, it inhibits amine
coordination, and this results in a lower reaction rate.
Equation (3), and this confirms the obtained first-order for
the palladium concentration (Figure 4).
v ˆ k[1][NaOtBu]
(3)
The mechanism depicted in Scheme 5 gives, according to
Equation (2), a lower reaction rate at higher bromide
concentrations. With cis chelating diphosphane Pd complexes,
prepared in situ, the amination of aryl triflates was reported to
be inhibited by added halides.[12]
Table 3 shows, however, that in this study halide salts
accelerate the reaction dramatically. This effect levels off at
higher halide concentrations. At even higher halide salt
concentrations, the reaction is somewhat inhibited. These
results can be explained in terms of two effects: an inhibitive
effect as expected from Equation (2), and an accelerating
effect due to the increased polarity of the solvent. Sodium tert-
butoxide is likely to exist as a tight ion pair in toluene.[17]
Addition of a salt could disrupt this ion pair, and this leaves a
more activated butoxide. This effect is comparable to the rate
enhancement obtained with addition of crown ether[18] and
the higher activity of sodium tert-butoxide compared with
lithium tert-butoxide.[19] The butoxide is activated and reacts
with the cationic species to form the butoxide complex.
Furthermore, addition of a salt increases the solvent polarity.
This more polar solvent might shift the equilibrium towards
the reactive, cationic species 2.
The NMR spectra under catalytic conditions show the
presence of starting material 1e (d ˆ 9.4), in exchange with
cationic species (2e). These results demonstrate that com-
plexes 2 are true intermediates in the catalytic cycle. The
reaction product of [Pd(PP)(Ar)(Br)] with sodium tert-
butoxide, probably intermediate [Pd(PP)(Ar)(OtBu)] (3),
could not be observed under catalytic conditions. This is as
expected when this compound is formed in the rate-limiting
step.
With the monodentate tris(o-tolyl)phosphane ligand, a
neutral Pd(amine) complex (viz. [Pd(P)(Br)(Ar)(amine)])
analogous to cationic complex 4 has been observed.[13] Coor-
ꢀ
dination to palladium enhances the N H acidity and after
deprotonation, the anionic compound [Pd(P)(Br)(Ar)-
(amido)]ꢀ was formed.[15]
Catalysis with neutral complexes (1): In contrast to the
cationic complexes 2, the use of neutral complexes 1 in the
amination of bromobenzene shows that the reaction rate is
independent of the amine concentration. From this we might
conclude that the reaction follows a different pathway.
Furthermore, at low conversion and thus low bromide
concentration, the reaction rate shows a dependency on
the palladium concentration which is of broken order (see
Figure 3). At higher bromide concentrations, first-order
dependence on the palladium concentration is recovered
(see Figure 4).
High halide concentration: The kinetic data can be explained
by the reaction sequence depicted in Scheme 5, if we consider
a fast equilibrium that involves decoordination of the
bromide. This equilibrium precedes the addition of the base,
which is the rate-limiting step (k2), as shown in Scheme 5.
Buchwald and co-workers have reported a series of cis-
[Pd(PP)(OR)(Ar)] analogues of 3, and this shows that these
compounds are stable.[16]
Low halide concentration: At low bromide concentration, the
reaction rate shows a positive, broken, order in [Pd], a first
order in sodium tert-butoxide, and a zeroth order in both
2-methoxyaniline and bromobenzene. The observed depen-
dence on the palladium concentration might suggest that
multiple equilibria, for example, between a dimeric species as
the resting state and a monomeric species, which is the active
intermediate, exist. From NMR spectroscopy, however, we
can conclude that probably Pd dimers are not involved as the
resting state, since under catalytic conditions only complex-
es 1e and 2e are observed.
As can be seen from this rate law, the halide formed in the
initial stage of the reaction will retard the rate of catalysis.
This decelerating effect will be more pronounced for more
active systems, which generate free halide at a higher initial
rate. Thus, experiments with high Pd loadings will give lower
TOFinis, and therefore a broken order in [Pd] will be observed.
At the same time, an increasing halide concentration will
activate the base, as explained above. Therefore, at low halide
concentrations opposing rate effects exist, which obscure the
kinetics.
+
P
P Pd
Br
P
O Pd
P
P
k2
k1
P Pd
Product
k-1
OtBu
1
2
3
Scheme 5. Proposed equilibrium between neutral compounds 1 and cat-
ionic compounds 2 in the amination of aryl bromides.
This dissociative mechanism gives the rate law expressed in
Equation (2).
k1k2‰1Š‰NaOtBuŠ
v ˆ
(2)
-
kꢀ1‰Br Š
To verify the reaction sequence of Scheme 5, we carried out
kinetic experiments with 1e as the catalyst in the presence of
50 equivalents of tetrabutylammonium bromide (with respect
to the palladium concentration). Under these conditions, the
rate law given in Equation (2) can be expressed as in
480
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0480 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 2

Palladium-Catalyzed Amination of Aryl Bromides and Aryl Triflates
475±482
bromobenzene (1.0 mmol) or phenyl triflate (1.0 mmol). When appropri-
ate, tetrabutyl ammonium bromide, tetrabutyl ammonium iodide, or
sodium chloride was added. The temperature was kept at 258C. Then,
NaOtBu (135 mg, 1.4 mmol) was added (t ˆ 0). The reaction was monitored
by taking samples from the reaction mixture at specific intervals. Each
sample was diluted with diethylether, washed with water, and the resulting
organic fraction was analyzed by gas chromatography.
Additionally, at low halide concentration the reaction might
also proceed by the alternative, cationic pathway (vide supra).
As a result, the two routes can be operative simultaneously,
and this results in complicated (i.e. nonunity order) kinetics
with respect to [Pd]. However, since no dependence on amine
concentration is observed, the role of the latter possibility is
uncertain.
Kinetic studies: The kinetic studies were performed under the same
conditions as described for the catalysis experiments, and the initial
concentrations of one reagent were varied at a time. The total volume of
the reaction was kept constant at 2.5 mL by adjusting the amount of
toluene added.
Bite angle effect: At low bromide concentrations, the initial
ꢀ
ꢀ
reaction is faster for compounds with a larger P Pd P angle.
This is due to the closer proximity of the oxygen to the Pd
center; this assists the dissociation of the bromide and thus
forces the equilibrium more to the active compounds 2.
Therefore, a higher concentration of complex 2 results in a
higher overall reaction rate.
NMR experiment: A frozen solution of complex 1e (20 mg, 0.023 mmol),
bromobenzene (157 mg, 1.0 mmol), 2-methoxyaniline (148 mg, 1.2 mmol),
and sodium tert-butoxide (55.5 mg, 0.6 mmol) in [D8]toluene (2 mL) was
allowed to melt before introduction in the NMR probe.
Acknowledgements
Conclusion
We thank the Netherlands Ministry of Economic Affairs (IOP Katalyse)
for support for this research.
A series of cationic and neutral [Pd(PP)(Ar)] complexes has
been synthesized and used as catalysts in the amination of aryl
triflates and aryl bromides. Kinetic studies show that for the
cationic complexes, the rate-limiting step is the deprotonation
of an Pd(amine) complex to form an amido complex. The
amine complex is formed in a fast equilibrium from the
starting cationic complex. In this reaction, the reaction rate is
enhanced by smaller bite angles.
In the presence of halide salts, the sodium tert-butoxide is
activated, and this results in a change of the reaction. The
alternative pathway proceeds by the rate-determining forma-
tion of the Pd(alkoxide) complexes 3. At higher concentra-
tions of halides, the rate-enhancing effect of the salt levels off.
This is due to a shift of the fast equilibrium between the active
cationic species 2 and the inactive neutral compound 1
towards the latter, inactive species.
[1] a) J. F. Hartwig, Angew. Chem. 1998, 110, 2154 ± 2177; Angew. Chem.
Int. Ed. 1998, 37, 2046 ± 2067; b) J. F. Hartwig, Synlett. 1997, 329 ± 340;
c) J. P. Wolfe, S. Wagaw, J.-F. Marcoux, S. L. Buchwald, Acc. Chem.
Res. 1998, 31, 805 ± 818; d) B. H. Yang, S. L. Buchwald, J. Organomet.
Chem. 1999, 576, 125 ± 146.
[2] a) B. C. Hamann, J. F. Hartwig, J. Am. Chem. Soc. 1998, 120, 7369 ±
7370; b) J. P. Sadighi, M. C. Harris, S. L. Buchwald, Tetrahedron Lett.
1998, 39, 5327 ± 5330.
[3] D. W. Old, J. P. Wolfe, S. L. Buchwald, J. Am. Chem. Soc. 1998, 120,
9722 ± 9723.
[4] J. P. Wolfe, S. L. Buchwald, J. Org. Chem. 2000, 65, 1157 ± 1167.
[5] J. F. Hartwig, Acc. Chem. Res. 1998, 31, 852 ± 860; b) G. Mann, J. F.
Hartwig, J. Am. Chem. Soc. 1996, 118, 13109 ± 13110.
[6] a) J. P. Wolfe, S. Wagaw, S. L. Buchwald, J. Am. Chem. Soc. 1996, 118,
7215 ± 7216; b) J. P. Wolfe, S. L. Buchwald, Tetrahedron Lett. 1997, 38,
5176 ± 5185; c) J. P. Wolfe, S. L. Buchwald, J. Org. Chem. 2000, 65,
1144 ± 1157.
[7] Y. Guari, D. S. van Es, J. N. H. Reek, P. C. J. Kamer, P. W. N. M.
van Leeuwen, Tetrahedron Lett. 1999, 40, 3789 ± 3790.
[8] M. A. Zuideveld, B. H. G. Swennenhuis, M. D. K. Boele, Y. Guari,
J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van Leeuwen, K. Goubitz, J.
Fraanje, M. Lutz, A. L. Spek, G. P. F. van Strijdonck, unpublished
results.
The higher rates obtained for complexes that contain
ligands with larger bite angles can be correlated with a higher
concentration of the corresponding active species 2. Such an
effect can be attributed to an additional driving force due to
the ability of Xantphos to act as a terdentate ligand.
[9] Reported X-ray crystal structures for complexes of the type
ꢀ
ꢀ
[Pd(P P)(X)(Ar)] with P P ˆ chelating phosphane, X ˆ halide, and
Arˆ aromatic group: a) V. Dufaud, J. Thivolle-Cazat, J.-M. Basset, R.
Mathieu, J. Jaud, J. Waisserman, Organometallics 1991, 10, 4005 ±
4015; b) A. Bohm, K. Polborn, K. Sunkel, W. Beck, Z. Naturforsch.
B 1998, 53, 448; c) W. A. Herrmann, C. Brossmer, T. H. Riermeier, K.
Öfele, J. Organomet. Chem. 1994, 481, 97 ± 108; d) I. R. Butler, L. J.
Hobson, S. J. Coles, M. B. Hurthouse, K. M. Abdul Malik, J. Organo-
met. Chem. 1997, 540, 27 ± 40; e) H. Trauner, P. le Floch, J.-M. Lefour,
L. Ricard, F. Mathey, Synthesis 1995, 717 ± 726; f) J. M. Brown, J. J.
Perez-Torrente, N. W. Alcock, H. J. Clase, Organometallics 1995, 14,
207 ± 213.
Experimental Section
General remarks: All reactions were carried out under
a nitrogen
atmosphere using standard Schlenk techniques. All solvents were freshly
distilled from standard drying agents under nitrogen before use. Phenyl
triflate, bromobenzene, and o-anisidine were distilled before use. Sodium
tert-butoxide, tetrabutyl ammonium bromide, tetrabutyl ammonium iodide,
and sodium chloride were used as received.
31P NMR spectra (121.5 MHz, external 85% H3PO4, CDCl3) were recorded
with a Bruker AMX-300 spectrometer. The product distribution was
measured with an Interscience Mega2 apparatus equipped with a DB1
column (length 30 m, inner diameter 0.32 mm, film thickness 3.0 mm) and
an FID detector.
[10] C. Amatore, G. Broeker, A. Jutand, F. Khalil, J. Am. Chem. Soc. 1997,
119, 7215 ± 7216.
[11] P. W. N. M. van Leeuwen, P. C. J. Kamer, J. N. H. Reek, P. Dierkes,
Chem. Rev. 2000, 100, 2741 ± 2770.
[12] J. Louie, M. S. Driver, B. C. Hamann, J. F. Hartwig, J. Org. Chem. 1997,
62, 1268 ± 1273.
The preparation and full characterization of the complexes will be
described elsewhere.[8]
Catalysis: In a typical experiment, a Schlenk vessel, fitted with a septum
[13] a) M. S. Driver, J. F. Hartwig, J. Am. Chem. Soc. 1997, 119, 8232 ± 8245;
b) F. Paul, J. Patt, J. F. Hartwig, Organometallics 1995, 14, 3030 ± 3039;
c) R. A. Widenhoefer, S. L. Buchwald, Organometallics 1996, 15,
3534 ± 3542; d) R. A. Widenhoefer, S. L. Buchwald, Organometallics
and
a magnetic stirring bar, was charged under nitrogen with the
appropriate complex (0.0049 mmol), toluene (2 mL), decane (104 mL,
1 mmol) as internal standard, 2-methoxyaniline (135 mL, 1.2 mmol), and
Chem. Eur. J. 2001, 7, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0481 $ 17.50+.50/0
481

P. W. N. M. van Leeuwen et al.
FULL PAPER
1996, 15, 2755 ± 2763; e) R. A. Widenhoefer, H. A. Zhong, S. L.
Buchwald, Organometallics 1996, 15, 2745 ± 2754.
[14] G. P. F. van Strijdonck, M. D. K. Boele, P. C. J. Kamer, J. G. de Vries,
P. W. N. M. van Leeuwen, Eur. J. Inorg. Chem. 1999, 1073 ±
1076.
[16] R. A. Widenhoefer, H. A. Zhong, S. L. Buchwald, J. Am. Chem. Soc.
1997, 119, 6787 ± 6795.
[17] A. Jutand, A. Mosleh, Organometallics 1995, 14, 1810 ± 1817.
[18] J. P. Wolfe, S. L. Buchwald, J. Org. Chem. 1997, 62, 6066 ± 6068.
[19] J. Louie, J. F. Hartwig, Tetrahedron Lett. 1995, 3609 ± 3612.
[15] J. Louie, F. Paul, J. F. Hartwig, Organometallics 1996, 15, 2794 ±
3005.
Received: March 22, 2000 [F2380]
482
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0702-0482 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 2