Palladium catalyst systems for cross-coupling reactions of aryl chlorides and olefins

Article abstract of DOI:10.1002/1521-3765(20010702)7:13<2908::AID-CHEM2908>3.0.CO;2-R

A detailed investigation into the influence of phosphines, additives, bases and solvents on the Heck coupling reaction of 4-trifluoromethyl-1-chlorobenzene (2) is presented. It is shown that a number of catalyst systems exist for efficient cross coupling of electron-deficient aryl chlorides with various olefins. Basicity and steric demand of the ligand are two factors which determine the success of the reaction. In addition the phosphine/palladium ratio, the correct type and amount of additive, and finally the use of an appropriate base and solvent are also important. The optimised reaction conditions are applied for the arylation of styrene, 2-ethylhexyl acrylate and N,N-dimethyl acrylic amide with various aryl chlorides.

Full text of DOI:10.1002/1521-3765(20010702)7:13<2908::AID-CHEM2908>3.0.CO;2-R

FULL PAPER
Palladium Catalyst Systems for Cross-Coupling Reactions of Aryl Chlorides
and Olefins**
Alexander Zapf and Matthias Beller*[a]
Abstract: A detailed investigation into
the influence of phosphines, additives,
bases and solvents on the Heck coupling
reaction of 4-trifluoromethyl-1-chloro-
benzene (2) is presented. It is shown that
a number of catalyst systems exist for
efficient cross coupling of electron-defi-
cient aryl chlorides with various olefins.
Basicity and steric demand of the ligand
are two factors which determine the
success of the reaction. In addition the
phosphine/palladium ratio, the correct
type and amount of additive, and finally
the use of an appropriate base and
solvent are also important. The opti-
mised reaction conditions are applied
for the arylation of styrene, 2-ethylhexyl
acrylate and N,N-dimethyl acrylic amide
with various aryl chlorides.
Keywords: additives ´ arylation ´
C Cl activation ´ homogeneous cat-
alysis ´ palladium
palladium(ii) acetate/phosphine in situ systems in the coupling
reaction of 4-trifluoromethyl-1-chlorobenzene (2) and styrene
(3). The optimised reaction conditions were applied to the
Heck reaction of a variety of aryl chlorides with different
olefins.
Introduction
Among the various ways known to synthesise arylated olefins,
probably the most powerful method is the construction of
C(sp2) C(sp2) single bonds through the palladium-catalysed
coupling reactions of olefins with aryl halides (Heck reac-
tion).[1] In general, aryl bromides, iodides and triflates are
used as starting materials for the Heck reaction. Due to the
industrial importance of being able to functionalise easily
accessible and cheap aryl chlorides,[2] there is currently a great
deal of interest in the coupling of aryl chlorides with various
nucleophiles.[3] While Bozell[4] demonstrated that nickel/
palladium mixtures promote effective coupling reactions with
olefins, most studies have focussed on palladium catalysts.
Although initial work conducted by W. A. Herrmann and us,[5]
as well as D. Milstein,[6] and more recently by M. T. Reetz,[7]
G. Fu,[8] and J. F. Hartwig[9] led to significant breakthroughs in
this area, the factors that affect the efficiency of the coupling
of aryl chlorides with olefins are far from being understood. It
is generally accepted that certain recipes for ªgoodº reaction
conditions exist, but the exact influence of most reaction
parameters remains unclear. In this full paper we report in
detail the effect of additives, bases and solvents on the
catalytic performance of the palladacycle 1, and of various
Results and Discussion
Initially, we performed the coupling of 4-trifluoromethyl-1-
chlorobenzene (2) and styrene (3) in the presence of pallada-
cycle 1 as a model reaction (Scheme 1). Here, we studied the
(o-tol)2
P
O
O
O
Pd
Pd
O
P
(o-tol)2
1
Cl
"Pd"
+
F3C
F3C
Scheme 1. Heck reaction of 2 and 3.
effect of added halide (cations and anions), bases and the
olefin concentration on catalyst productivity. All reactions
were run at a low catalyst concentration (0.05 mol% 1 ˆ
0.1 mol% Pd). The negative inductive effect of the para-
trifluoromethyl substituent leads to an activation of the C Cl
bond towards nucleophilic substitution.[10] Hence, compound
2 should be viewed as a moderately activated aryl chloride for
palladium-catalysed coupling reactions.
[a] Prof. Dr. M. Beller, Dr. A. Zapf
Institut für Organische Katalyseforschung
an der Universtität Rostock e. V. (IfOK)
Buchbinderstr. 5 ± 6, 18055 Rostock (Germany)
Fax : (‡49)381-4669324
E-mail: matthias.beller@ifok.uni-rostock.de
[**] Palladium-Catalyzed Reactions for Fine Chemical Synthesis, Part 20;
Â
for Part 19 see M. Gomez Andreu, A. Zapf, M. Beller, Chem.
Commun. 2000, 2475 ± 2476.
2908
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2908 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 13

2908±2915
In order to achieve significant conversion of chloroarenes in
the palladacycle-catalysed Heck reaction comparatively high
reaction temperatures (>1408C) are needed owing to the
relative stability of the C Cl bond towards oxidative addi-
tion.[2a] Only the highly basic catalyst systems developed very
recently do not need this thermal activation and allow Heck
coupling of aryl chlorides at 100 ± 1208C.[6, 8, 9]
The palladacycle 1 is an ideal source for the slow generation
of palladium monophosphine species, which are most likely
the active catalytic species in
presence of 5 ± 50 mol% TBAB. These results demonstrate
that TBAB has both a positive and negative effect on the yield
of 4 depending on the concentration and temperature. This
can be explained by the fact that bromide ions facilitate the
formation of anionic 14e and 16e Pd0 monophosphine
species (Scheme 2). At 1608C the formation of these highly
active species is fast and a higher halide concentration leads to
the predominant formation of stable 18e complexes. At
1408C the conversion of 1 to anionic Pd0 complexes is low,
the reaction mixture.[5] In or-
der to prevent deactivation
(precipitation of palladium
black) of these active species
at higher reaction tempera-
tures, it is necessary to add
stabilising and/or activating
agents to the reaction mixture.
Soluble sources of bromide
ions often show a beneficial
effect in Heck reactions (Jeff-
ery conditions).[1] Hence, we
Scheme 2. Proposed equilibrium of Pd complexes (for simplification there is no equilibrium with 12e and 14e
Pd species shown).
studied the influence of different concentrations of tetra-n-
butylammonium bromide (TBAB) on the model reaction at
1408C and 1608C. Without any TBAB in the presence of
0.05 mol% 1 and 1.2 equivalents of sodium acetate as the base
only 2% and 8% of 4-trifluoromethyl stilbene (4) is formed
after 20 hours at 1408C and 1608C, respectively. Increasing
the halide concentration leads to a higher product yield at
both temperatures. However, there are significant differences
depending on the reaction temperature. As shown in Figure 1
hence a higher halide concentration is beneficial for the initial
step. Nevertheless, reactivity of the 16e Pd0 species is not
high enough at 1408C to ensure sufficient conversion of the
test substrate.
While the initial catalyst tests were performed with sodium
acetate as the base, we were also interested in the influence of
other bases. Common bases for Heck reactions are the
acetates, carbonates and bicarbonates of alkaline or earth
alkaline metals, or alternatively organic bases such as trialkyl-
amines.[2] These bases may also act as ligands at the Pd centre,
thereby influencing the catalyst performance. Table 1 gives an
overview of the results obtained in the presence of 1.2 equiv-
alents of ten different bases. Sodium and potassium acetate
and sodium carbonate lead to similar results in the presence of
20 mol% TBAB (64, 61 and 61% of 4, respectively). Soluble
tetra-n-butylammonium acetate leads, without added halide,
to the formation of 32% of 4. This is surprising because
acetate ions should have a comparable stabilising effect on
palladium(0) species as bromide ions due to ate-complex
formation.[11] The expensive base cesium carbonate (25%) as
Figure 1. Effect of TBAB concentration (10 mmol 2, 15 mmol 3, 12 mmol
sodium acetate, 0.005 mmol 1, 10 mL DMAc, 20 h).
Table 1. Base effect on the coupling of 2 and 3.[a]
Base
Bu4NBr [mol%]
Yield [%][b]
at 1608C the optimum concentration of TBAB lies at
20 mol% relative to 2, with 64% of 4 being formed (turnover
number (TON) ˆ 640). At higher concentrations of TBAB the
yield decreases, with only 10% of 4 being obtained using
100 mol% TBAB. We explain this decrease in product yield at
higher concentrations of TBAB by the blocking of free
coordination sites of low-ligated palladium(0) complexes.
At 1408C the yield of 4 is significantly lower compared to
1608C; this demonstrates that a temperature of 1408C is too
low for an efficient activation of this substrate. After an initial
increase in the yield by adding 5 mol% of TBAB the halide
concentration has only a marginal influence on the reaction
outcome. Hence, 4 is obtained in about 20% yield in the
1
2
3
4
5
6
7
8
9
NaOAc
KOAc
Bu4NOAc
Na2CO3
Cs2CO3
K3PO4
KOtBu
NaOH
CaO
20
20
0
20
20
20
20
20
20
20
64
61
32
61
25
39
3
4
67
12
10
NEt3
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol base, 0.005 mmol
1, 10 mL DMAc, 1608C, 20 h. [b] Yield of 4, determined by gas
chromatography with diethyleneglycol di-n-butyl ether as the internal
standard.
Chem. Eur. J. 2001, 7, No. 13
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2909 $ 17.50+.50/0
2909

M. Beller, A. Zapf
FULL PAPER
well as potassium phosphate (39%) gave lower yields of 4.
Both potassium tert-butoxide and sodium hydroxide lead to
catalyst decomposition under these conditions. Much to our
surprise, simple calcium oxide, almost insoluble in N,N-
dimethyl acetamide (DMAc) and to the best of our knowl-
edge not used before in palladium-catalysed coupling reac-
tions, performs excellently under these reaction conditions
(67% of 4). Triethylamine is not suitable as base (12% of 4)
due to its ability to bind strongly to palladium. In all catalytic
tests the selectivity for the E isomer of 4 lies between 94 and
98%, independent of the type of based used.
From these results we conclude that it is important to have a
low concentration of basic anions in solution. Again the
blocking of free coordination sites on the metal centre, which
is necessary for the turnover of the catalyst, might be the
reason for this effect.
After considerable optimisation of the model reaction, we
were interested in the outcome of this coupling reaction in the
presence of a series of halides and other coordinating and
non-coordinating salts. Investigations into the oxidative
addition of aryl iodides to palladium(0) complexes confirm
an activating and, hence, accelerating effect of added halide
ions on this step of the catalytic cycle.[12] With regard to the
results obtained with TBAB as the additive we believe that
this is also valid for the coupling of aryl chlorides. Due to the
fact that the addition of other anions (chloride,[13] iodide,[14]
hydrogen sulfate,[15] acetate[16]) to palladium-catalysed cou-
pling reactions is also beneficial in some cases, we tested
various additives in the presence of sodium acetate and
sodium carbonate (Tables 2 and 3). Table 2 gives a summary
of the results observed in the presence of different tetra-n-
butylammonium salts. As can be seen, bromide is the most
effective counterion (TBAB) for this reaction if sodium
acetate is used as the base. When sodium carbonate is utilised
as the base, better results are obtained in the presence of
tetra-n-butylammonium chloride hydrate (89% of 4), fol-
lowed then by the corresponding bromide and hydrogen
sulfate (61 and 58% of 4, respectively). By reaction with
carbonate ions, the chlorideꢁs crystal water can form hydrox-
ide ions that are able to coordinate to palladium and, hence,
lead to considerable stabilisation of the active species. This
formation of hydroxide ions seems to be responsible for the
beneficial effect of chloride and water compared to chloride
without water. The basicity of acetate ions is not sufficient to
deprotonate water to a reasonable extent. In this case, the
combination of water and chloride has no significant positive
influence on the catalytic efficiency. Weakly coordinating
anions (acetate, nitrate) as well as non-coordinating anions
(tetrafluoroborate, sulfonic acid anions) do not stabilise low
ligated palladium(0) complexes towards agglomeration and,
hence, do not allow reasonable conversion of the test substrate.
During the investigations into the anionꢁs influence on
catalytic activity it became evident that the cation might also
have a significant impact on the success of the reaction. The
results of screening different halide additives are summarised
in Table 3. Readily soluble tetraalkylammonium and lithium
Interestingly, the ratio of aryl chloride to olefin is crucial for
the outcome of the coupling reaction (Figure 2). A yield of
91% of the desired stilbene 4 (relative to styrene) is observed
Figure 2. Influence of the styrene/aryl chloride ratio (10 mmol 2, 2 mmol
TBAB, 12 mmol sodium acetate, 0.005 mmol 1, 10 mL DMAc, 1608C,
20 h).
by utilising 0.5 equivalents of styrene relative to 2. Equimolar
amounts of starting materials lead to a 66% yield of 4. A
further increase in the styrene/aryl chloride ratio (1.5± 5 equiv)
results in a continuous decrease in the product yield
(64 ± 23%).
Table 2. Anion effect on the coupling of 2 and 3.[a]
Despite the lower yields, we
decided to use 1 ± 1.5 equiva-
lents of the olefin throughout
our study because this is the
cheaper coupling partner rela-
tive to the aryl halide. The
effect of olefin concentration
on the catalyst performance is
extremely important for an ob-
jective comparison of different
catalyst systems. From the re-
sults shown above, we conclude
that one can often influence the
catalyst activity and productiv-
ity more by changing the con-
centration of olefin than by
switching from one catalyst to
another.
Cation
Anion
NaOAc
Yield [%][b]
Select. E [%]
Na2CO3
Yield [%][b]
Select. E [%]
‡
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Bu4N
Cl
21
36
64
35
1
97
96
94
91
nd[c]
nd
95
96
±
nd
97
nd
nd
nd
29
89
61
11
0
98
95
94
94
±
‡
Cl ´ H2O
Br
I
SCN
SCN (0.1 mol%)
OAc
‡
‡
‡
‡
1
1
nd
95
96
±
‡
34
23
0
5
16
2
43
21
0
‡
OH (H2O solvent)
OH (MeOH solvent)
‡
‡
NO3
HSO4
BF4
OTs
OTf
9
58
0
nd
95
±
‡
‡
‡
4
1
10
12
nd
nd
‡
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol base, 2 mmol additive, 0.005 mmol 1, 10 mL DMAc,
1608C, 20 h. [b] Yield of 4, determined by gas chromatography with diethyleneglycol di-n-butyl ether as the
internal standard. [c] Not determined.
2910
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2910 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 13

Palladium Catalyst Systems
2908±2915
Table 3. Cation effect on the coupling of 2 and 3.[a]
the presence of the economi-
cally attractive base sodium
carbonate an 85 ± 90% yield
(TON ˆ 850 ± 900) of 4 was re-
alised by applying either
0.2 equivalents tetra-n-butyl-
phosphonium bromide, tet-
ra(N,N-diethylamino)phospho-
nium bromide or tetra-n-butyl-
ammonium chloride hydrate.
Next, we were interested in a
comparison of the molecularly
defined complex 1 with tradi-
tional in situ catalyst systems.
As stated previously palladi-
Cation
Anion
NaOAc
Select. E [%]
Na2CO3
Select. E [%]
Yield [%][b]
Yield [%][b]
‡
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Li
Cl
Cl
Cl
15
8
97
nd
97
96
93
94
94
95
95
94
93
93
95
90
91
91
10
13
29
89
21
19
53
28
30
61
67
90
85
8
nd[c]
nd
98
95
95
95
95
96
97
94
93
93
94
94
nd
94
‡
Cs
‡
Bu4N
21
36
54
24
29
22
31
64
56
60
22
39
33
35
‡
Bu4N
Cl ´ H2O
‡
Li
Br
Br
Br
Br
Br
Br
Br
Br
Br
I
‡
Na
‡
K
‡
NH4
Et4N
Bu4N
Oc4N
Bu4P
(Et2N)4P
Me4N
Et4N
‡
‡
‡
‡
‡
‡
‡
I
I
10
11
um(ii)/phosphine
mixtures,
‡
Bu4N
which are generally applied for
Heck and related reactions
were not tested systematically
for the activation of aryl chlor-
ides at low catalyst loading
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol base, 2 mmol additive, 0.005 mmol 1, 10 mL DMAc,
1608C, 20 h. [b] Yield of 4, determined by gas chromatography with diethyleneglycol di-n-butyl ether as internal
the standard. [c] Not determined.
bromides form the most active catalysts in combination with
sodium acetate, while less soluble alkaline bromides lead to
low conversions. Utilising sodium carbonate the Hofmann
degradation of tetra-n-butylphosphonium bromide to form
tri-n-butylphosphine, which leads to an active catalyst system
itself, enhances the effect of simply added bromide. The
reasons for the different results combining tetra(N,N-dieth-
ylamino)phosphonium bromide with sodium acetate and
carbonate are not clear at the moment.
At this time the cationic effect of additives seems to depend
solely on their solubility. However, other ways of interacting
with the substrate or the catalyst system cannot be ruled out
with certainty. Therefore we performed NMR experiments to
detect an interaction between the added cation and the
chlorine of the aryl chloride and to observe an interaction of
the cation with palladium phosphine species. A change in the
chemical shift of 13C signal of the ipso-carbon atom could not
be observed in DMAc/[D6]DMSO (5:1) at room temperature
(d ˆ 137.2 ± 137.6). Also an influence of the cation on the
chemical shift of triphenylphosphine/palladium(0)/bromide
ate-complexes could not be verified. In the 31P NMR under
reaction conditions, albeit at room temperature, two signals
(0.1 mol% Pd) prior to our investigations. In order to observe
positive as well as negative influences of the different
commercially available phosphine ligands, we performed the
model reaction under standard reaction conditions (DMAc,
1608C, 20 h) with tetra-n-butylammonium bromide as the
additive. Typically, phosphines are applied in palladium-
catalysed coupling reactions in a molar ratio of 1:1 to 4:1 with
respect to palladium. It is said that higher phosphine/
palladium ratios slow down the coupling reaction.[18] There-
fore we chose a P/Pd ratio of 2:1 for the phosphine screening.
As shown in Table 4 most of the investigated ligands do not
catalyse the test reaction to an appreciable extent with or
without 20 mol% of tetra-n-butylammonium bromide.
In each case less than 10% of desired stilbene 4 was formed
in the absence of TBAB. Only two phosphines led to
significantly active catalyst systems under these conditions:
tricyclohexylphosphine (71% yield of 4 in the presence of
sodium carbonate) and tri-n-butylphosphine (83% yield of 4
in the presence of sodium carbonate), which are among the
most basic ligands investigated in this study. However, basicity
is not the only important factor that determines the success of
the reaction: tris(dimethylamino)phosphine, which is only a
bit less basic than tri-n-butylphosphine, is not suitable for the
model reaction at all. On the other hand steric effects are also
not the sole determining factor, for example, tricyclohexyl-
phosphine and tri-n-butylphosphine work almost equally well.
More electron-poor ligands whose size is comparable to the
former [tris(perfluorophenyl)phosphine] or the latter [tris(2-
cyanoethyl)phosphine] ligand do not support the reaction
(<10% yield of 4). Tri-o-tolylphosphine does not lead to an
active catalyst system when applied in situ. Here, palladium
precipitation occurs faster than complexation and concom-
itant cyclometallation. Surprisingly, tri-tert-butylphosphine/
palladium(ii) acetate does not catalyse the coupling of our test
substrates under these conditions, although it is known to
generate a highly efficient catalysts system in etheral solvents
with cesium carbonate as a base.[8] In all reactions described
here the selectivity towards the E double bond isomer lies
are observed for
a mixture of Pd(PPh3)4/Br /DMAc/
[D6]DMSO (5:1). Both signals are broad and an assignment
is difficult because the two signals represent a variety of both
complexes and free phosphine that are all in fast equilibrium.
As shown in Tables 2 and 3 the additive did not significantly
influence the geometry of the double bond formed (E vs Z).
In all cases 93 ± 97% of the E isomer was obtained. However,
iodide salts seem to decrease the selectivity marginally (ca.
91% E) relative to the other anions investigated.
In the presence of 20 mol% of potassium trichlorostannate
or various other transition metal halides as additives, for
example, manganese(ii), cobalt(ii), nickel(ii) (compare ref. [4])
or copper(i), no efficient conversion of 2 was observed (less
than 20% yield of 4).
At this point three efficient catalyst systems based on the
palladacycle 1 for the activation of 2 have been discovered. In
Chem. Eur. J. 2001, 7, No. 13
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2911 $ 17.50+.50/0
2911

M. Beller, A. Zapf
FULL PAPER
Table 4. Variation of phosphine ligands (P/Pd(OAc)2 ˆ 2:1) in the coupling of 2 and 3.[a]
reactions and may be substitut-
ed by each other. Higher con-
centrations of phosphine ligand
lead to a decrease in catalyst
activity (here the blocking of
free coordination sites domi-
nates), although the catalyst is
stabilised against decomposi-
tion. Interestingly, increasing
the phosphine/palladium ratio
results in a significant decrease
of E selectivity in the coupling
product (P/Pd ˆ 10:1, E selec-
tivity of 4 ˆ 91%; P/Pd ˆ 100:1,
E selectivity of 4 ˆ 79%).
Phosphine
n
V
[8][c]
Bu4NBr
[mol%]
Yield
[%][d] (NaOAc)
Yield
1
[b]
[cm
]
[%][d] (Na2CO3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
PCy3
PCy3
P(tBu)3
P(nBu)3
2056
2056
2056
2060
2060
170
170
182
132
132
0
20
20
0
9
5
39 (94)
26 (95)
2
26 (96)
22
27 (95)
2
13
21 (96)
22 (95)
0
20
0
18 (96)
71 (94)
31 (94)
6
83 (94)
47 (95)
38 (95)
0
P(nBu)3
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
P[2,4,6-(MeO)3C6H2]3
PCy2(o-biph)
P(NMe2)3
PMe2Ph
dppb[e]
PMePh2
PPh2(2-Py)
dppf[f]
P(o-Tol)3
dppt[g]
PPh3
2062
2065
2066
2067
157
122
7
18 (96)
14
0
35
1
13
7
2
10
0
136
194
2067
The success in coupling our
test substrates in the presence
of an excess of triphenylphos-
phine prompted us to re-inves-
tigate the ligand influence at
higher phosphine/palladium ra-
tios (Table 5). All reactions re-
sulted in significantly higher
product yields when a tenfold
excess of ligand is applied com-
pared with the initial twofold
2069
2071
2078
2091
145
145
132
184
1
0
3
1
P(p-C6H4-F)3
P(CH2CH2CN)3
P(C6F5)3
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol base, 0 or 2 mmol TBAB, 0.01 mmol Pd(OAc)2, 10 mL
DMAc, 1608C, 20 h. [b] nCO(A1) of Ni(CO)3L in CH2Cl2.[17] [c] Tolman cone angle.[17] [d] Yield of 4 (selectivity of
E isomer in parentheses), determined by gas chromatography with diethyleneglycol di-n-butyl ether as the
internal standard. [e] 1,4-Bis(diphenylphosphino)butane. [f] 1,1'-Bis(diphenylphosphino)ferrocene. [g] 2,6-Bis-
(diphenylphosphino)toluene.
between 94 and 96%, similar to that found for the pallada-
cycle catalyst 1. Hence, the same type of palladium species,
which undergoes the b-hydride elimination, is likely.
In contrast to catalyst 1 the formation of palladium black is
observed in the presence of most of the shown ligands. Thus,
we assumed that the stabilisation of the active palladium
species by a slight excess of ligand (2:1) is not sufficient.
Therefore, the dependence of catalyst productivity on the
phosphine/palladium ratio was investigated. As shown in
Figure 3 the coupling of 2 with 3 can be affected efficiently by
excess. The only exception was tri-n-butylphosphine which led
to 53% of 4 compared with 83% at lower phosphine/
palladium ratio in the presence of sodium carbonate.
Some important conclusions can be drawn from these
results:
1) Regardless of the phosphine ligand employed, substantial
yields of Heck products can be obtained even in the
olefination of chloroarenes, if just the optimal conditions
with respect to P/Pd ratio, base and co-catalyst are used.
2) An important but generally overlooked reaction parame-
ter is the ligand/palladium ratio.
3) The combination of the right ligand/palladium ratio and
the reaction temperature allows an efficient activation of
C-Cl bonds at low catalyst loading.
Table 5. Variation of phosphine ligands (P/Pd(OAc)2 ˆ 10:1) in the
coupling of 2 and 3.[a]
Phosphine
Bu4NBr [mol%] Yield [%][b] Select. E [%]
1
2
3
4
5
6
7
8
9
PCy3
PCy3
P(tBu)3
P(nBu)3
P(nBu)3
P[2,4,6-(MeO)3C6H2]3 20
PCy2(o-biph)
dppb[d]
0
20
20
0
40
72
29
12
23
31
29
55
19
50
43
17
92
92
93
nd[c]
91
94
94
90
87
92
92
90
Figure 3. Influence of the P/Pd ratio (10 mmol 2, 15 mmol 3, 0 or 2 mmol
TBAB, 12 mmol sodium acetate, 0.01 mmol palladium(ii) acetate, 10 mL
DMAc, 1608C, 20 h).
20
20
20
20
20
20
20
PMePh2
means of simple triphenylphosphine if the ligand is provided
in an excess relative to palladium. The best yield of 4 (72%) is
observed at a phosphine/palladium ratio of 15:1 in the
presence of tetra-n-butylammonium bromide.
10 PAllPh2
11 dppf[e]
12 P(CH2CH2CN)3
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol NaOAc, 0 or
2 mmol TBAB, 0.01 mmol Pd(OAc)2, 10 mL DMAc, 1608C, 20 h. [b] Yield
of 4, determined by gas chromatography with diethyleneglycol di-n-butyl
ether as the internal standard. [c] Not determined. [d] 1,4-Bis(diphenyl-
phosphino)butane. [e] 1,1'-Bis(diphenylphosphino)ferrocene.
If the co-catalyst is omitted comparable results (65%) are
obtained at a slightly higher molar ratio of phosphine to
palladium (20:1). This again supports the proposal that
bromide ions and phosphines both act as ligands in the Heck
2912
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2912 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 13

Palladium Catalyst Systems
2908±2915
Remarkably, even catalyst systems composed of electron-
poor phosphite ligands efficiently catalyse the Heck coupling
of chloroarenes with various olefins, if the ligand is provided
in an excess.[19]
The last reaction parameter studied in the model reaction
was the influence of the solvent (Table 6). Heck reactions are
applied to the Heck reaction of various chloroarenes and
different olefins (Scheme 3; Table 7).
All substrates tested here were converted to the desired
coupling products in very good to excellent yields (85 ± 100%)
and with excellent turnover numbers (850 ± 1000) with respect
to Heck reactions of aryl chlorides. It is clear that for all
Table 6. Coupling of 2 and 3 in polar solvents.[a]
Solvent
Yield [%][b]
(NaOAc)
Yield [%][b]
(Na2CO3)
1
2
3
4
5
DMAc
NMP
TMU[c]
PEG 400[d]
ethylene glycol
56
63
48
25 (10)
11 (28)
40
59
48
2 (37)
2 (32)
Scheme 3. Coupling of aryl chlorides and olefins in the presence of
different catalyst systems.
[a] Reaction conditions: 10 mmol 2, 15 mmol 3, 12 mmol base, 0.01 mmol
Pd(OAc)2, 0.15 mmol PPh3, 9 mL solvent, 1 mL toluene, 1608C, 20 h.
[b] Yield of 4 (yield of trifluoromethylbenzene in parentheses), determined
by gas chromatography with diethyleneglycol di-n-butyl ether as the
internal standard. [c] N,N,N',N'-Tetramethylurea. [d] Polyethylene gly-
col 400.
shown combinations of substrates, more than one catalyst
system exists that leads to similar good results; ortho-
substituents at the aryl chloride are tolerated and electron-
poor olefins can be coupled as well as electron-rich ones.
Although in the latter case regio and stereochemical compli-
normally conducted in dipolar
Table 7. Coupling of aryl chlorides and different olefins.[a]
aprotic solvents like DMAc,
DMF or acetonitrile. Therefore
we investigated different polar
solvents such as N,N,N',N'-tet-
ramethylurea (TMU), N-meth-
ylpyrrolidone (NMP), DMAc,
polyethylene glycol and ethyl-
ene glycol in the presence of the
most economically attractive
catalyst system [Pd(OAc)2/
15PPh3]. Independent of the
base used, NMP gave slightly
better results with respect to
DMAc and TMU. In ethylene
glycol and polyethylene gly-
col 400 rapid palladium decom-
position takes place; hence,
these solvents do not allow an
efficient coupling of 2 and 3.
Reductive dehalogenation of 2
to trifluoromethylbenzene is
observed as a major side reac-
tion (up to 37%).
The presented studies of the
model reaction reveal that
there are far more than ten
different combinations of palla-
dium source, ligand, co-catalyst
(halide), base and P/Pd ratio
that are able to catalyse the
coupling of 2 and 3 in yields
from 60 ± 90% at a low catalyst
loading. Some of these ªbestº
catalyst systems were chosen and
ArCl
Olefin
Pd source
Ligand[b]
Bu4NBr [mol%]
20
Yield [%][c]
90
Cl
Cl
Cl
1
2
Pd(OAc)2
10P(OAr)3
F3C
N
Pd(OAc)2
10P(OAr)3
20
89
F3C
F3C
O
3
4
Pd(OAc)2
10P(OAr)3
±
20
93
94
Cl
[d]
1
CN
Cl
N
[d]
20
20
5
6
1
±
100
85
CN
O
Cl
Pd(OAc)2
2PBu3
O
O
Cl
Cl
N
7
8
Pd(OAc)2
Pd(OAc)2
10P(OAr)3
10PCy3
89
87
O
O
O
0
O
O
[a] Reaction conditions: 10 mmol ArCl, 15 mmol olefin, 12 mmol sodium carbonate, 0.01 mmol catalyst, 10 mL
DMAc, 1608C, 24 h. [b] Yield E product determined by gas chromatography with diethyleneglycol di-n-butyl
ether as the internal standard. [c] Arˆ 2,6-di-tert-butylphenyl. [d] 20 mol% Bu4NCl ´ H2O.
Chem. Eur. J. 2001, 7, No. 13
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2913 $ 17.50+.50/0
2913

M. Beller, A. Zapf
FULL PAPER
ether (400 mg, as the internal standard), an appropriate amount of additive
(if necessary), ligand and palladium source was suspended in dry DMAc
(10 mL) under an atmosphere of argon. The tube was sealed and put in a
preheated bath of silicon oil. After 20 hours the mixture was cooled to
room temperature and CH2Cl2 (10 mL) and HCl (2n, 10 mL) are added.
The organic phase was analysed by gas chromatography. After washing the
organic phase with water and brine, drying and evaporating the solvents,
the products were isolated by crystallisation from methanol/acetone
mixtures or by column chromatography (silica gel, hexane/ethyl acetate
mixtures).
cations arise from contrarily directing effects based on steric
and electronic demands, respectively, which were not further
investigated here. In the case of styrene, E/Z selectivity is
about 95/5 in all cases, whereas acrylic esters and amides are
coupled highly selectively (>99%) to form the desired E b-
product.
Utilising chlorobenzene and 3 as the substrates in the
presence of 0.1 mol% of palladium catalysts never gave more
than 13% of coupling product. Reaction of 4-chloroanisole
with styrene resulted in 17% of 4-methoxy stilbene in the
presence of palladium(ii) acetate/20triphenylphosphine. All
other catalyst systems under investigation led to lower yields
of the corresponding coupling product. Although the yields of
the Heck reaction of non-activated aryl chlorides are not
sufficient for organic synthesis, the achieved turnover num-
bers of 130 ± 170 are similar to the best previously described
catalysts with sterically hindered basic phosphines.
E-4-(Trifluoromethyl)stilbene: 1H NMR (360 MHz, CDCl3, 258C): d ˆ 7.60
(m, 4H), 7.54 (d, 3J(H,H) ˆ 7.3 Hz, 2H), 7.39 (m, 2H), 7.31 (m, 1H), 7.19 (d,
3J(H,H) ˆ 16.4 Hz, 1H), 7.11 (d, 3J(H,H) ˆ 16.4 Hz, 1H); 13C{1H} NMR
(90.6 MHz, CDCl3, 258C): d ˆ 140.8, 136.6, 131.2, 128.3, 127.1, 129.2 (q,
2J(C,F) ˆ 32.3 Hz), 128.8, 126.8, 126.5, 125.6 (q, 3J(C,F) ˆ 3.8 Hz), 124.2 (q,
‡
1J(C,F) ˆ 272 Hz); MS (70 eV, EI): m/z: 248 [M] , 179, 152, 89, 76.
N,N-Dimethyl-p-(trifluoromethyl)cinnamoyl amide: 1H NMR (360 MHz,
3
CDCl3, 258C): d ˆ 7.64 (d, J(H,H) ˆ 15.5 Hz, 1H), 7.58 (m, 4H), 6.93 (d,
3J(H,H) ˆ 15.5 Hz, 1H), 3.10 (s, 6H); 13C{1H} NMR (90.6 MHz, CDCl3,
2
258C): d ˆ 166.1, 140.6, 138.8, 131.0 (q, J(C,F) ˆ 32.4 Hz), 127.9, 125.7 (q,
1
3J(C,F) ˆ 3.4 Hz), 123.9 (q, J(C,F) ˆ 272 Hz), 120.0, 36.7; MS (70 eV, EI):
‡
m/z: 243 [M] , 224, 199, 171, 151, 98, 75.
E-3-(Trifluoromethyl)stilbene: 1H NMR (360 MHz, CDCl3, 258C): d ˆ 7.76
(m, 1H), 7.66 (m, 1H), 7.54 ± 7.44 (m, 4H), 7.39 (m, 2H), 7.31 (m, 1H), 7.17
(d, 3J(H,H) ˆ 16.4 Hz, 1H), 7.10 (d, 3J(H,H) ˆ 16.4 Hz, 1H); 13C{1H} NMR
(90.6 MHz, CDCl3, 258C): d ˆ 136.2, 137.5, 131.1 (q, 2J(C,F) ˆ 32.1 Hz),
130.6, 129.5, 129.1, 128.2, 127.1, 128.8, 126.7, 124.2 (q, 1J(C,F) ˆ 272 Hz),
124.0 (q, 3J(C,F) ˆ 3.8 Hz), 123.1 (q, 3J(C,F) ˆ 3.8 Hz); MS (70 eV, EI): m/z:
Conclusion
What can we learn from this study of various catalyst systems
for the Heck reaction of activated and non-activated aryl
chlorides and olefins? Firstly, for activated aryl chlorides a
number of useful catalyst systems exist that allow these
reactions to proceed in good to excellent yield with catalyst
turnover numbers of approximately 1000. The industrial
realisation of coupling reactions with these substrates in the
area of fine chemicals is nowadays, in our opinion, feasible.
Importantly, the outcome of the many possible combinations
of phosphine/additive/base is not easily predicted. While some
general guidelines exist, for example, to use a higher P/Pd
ratio, small changes in the co-catalyst or base may lead to a
dramatic decrease in catalyst productivity. Contrary to tradi-
tional belief, basicity or steric demand of the ligand is not
decisive for the success of the reaction. Here, the phosphine/
palladium ratio and the olefin concentration are more
important. In addition, one must note that the reaction
conditions once optimised for a special pair of coupling
partners are not automatically suitable for a similar coupling
reaction.
With regard to the Heck coupling of non-activated aryl
chlorides, more efficient catalyst systems are badly needed.
Although the catalyst productivity of the simple Pd/excess
PPh3 catalysts presented here is in the range of 100 ± 200,
product yields are low. However, even the new basic and
sterically demanding ligands such as tri-tert-butylphosphine[8]
or the new adamantylphosphines[3q, 20] allow the reaction to
proceed in good yield only at a comparable high catalyst
concentration (>1 mol% of Pd catalyst).
‡
248 [M] , 233, 179, 89.
2-Cyanostilbene: 1H NMR (400 MHz, CDCl3, 258C): d ˆ 7.76 (d,
2J(H,H) ˆ 8.1 Hz, 1H), 7.62 ± 7.52 (m, 4H), 7.44 ± 7.22 (m, 6H); 13C{1H}
NMR (101 MHz, CDCl3, 258C): d ˆ 140.4, 136.1, 133.3, 133.0, 132.7, 128.8,
‡
128.7, 127.5, 127.0, 125.2, 123.9, 117.9, 111.1; MS (70 eV, EI): m/z: 205 [M] ,
204, 190, 176, 102, 89.
N,N-Dimethyl-o-cyanocinnamoyl amide: 1H NMR (400 MHz, CDCl3,
258C): d ˆ 7.75 (d, 2J(H,H) ˆ 15.7 Hz, 1H), 7.64 ± 7.53 (m, 3H), 7.37 (m,
1H), 7.14 (d, 2J(H,H) ˆ 15.7 Hz, 1H), 3.13 (s, 3H), 3.02 (s, 3H); 13C{1H}
NMR (101 MHz, CDCl3, 258C): d ˆ 165.7, 138.3, 137.1, 133.6, 132.8, 129.1,
‡
128.2, 123.1, 117.7, 111.3, 37.4, 35.8; MS (70 eV, EI): m/z: 200 [M] , 156, 128,
98.
E-4-Acetylstilbene: 1H NMR (360 MHz, CDCl3, 258C): d ˆ 7.92 (d,
3J(H,H) ˆ 8.2 Hz, 2H), 7.54 (d, 3J(H,H) ˆ 8.2 Hz, 2H), 7.51 (d, 3J(H,H) ˆ
7.3 Hz, 2H), 7.34 (m, 3H), 7.19 (d, 3J(H,H) ˆ 16.3 Hz, 1H), 7.09 (d,
3J(H,H) ˆ 16.4 Hz, 1H), 2.56 (s, 3H); 13C{1H} NMR (90.6 MHz, CDCl3,
258C): d ˆ 197.3, 141.9, 136.7, 135.9, 131.4, 128.8, 128.7, 128.3, 127.4, 126.8,
‡
126.4, 26.5; MS (70 eV, EI): m/z: 222 [M] , 207, 178, 152, 89.
p-Acetyl-N,N-dimethylcinnamoyl amide: 1H NMR (360 MHz, CDCl3,
258C): d ˆ 7.87 (d, 3J(H,H) ˆ 8.0 Hz, 2H), 7.59 (d, 3J(H,H) ˆ 15.5 Hz,
3
3
1H), 7.52 (d, J(H,H) ˆ 8.0 Hz, 2H), 6.92 (d, J(H,H) ˆ 15.5 Hz, 1H), 3.12
(s, 3H), 3.00 (s, 3H), 2.52 (s, 3H); 13C{1H} NMR (90.6 MHz, CDCl3, 258C):
d ˆ 197.2, 165.9, 140.6, 119.9, 139.6, 137.3, 128.6, 127.7, 37.3, 35.8, 26.5; MS
‡
(70 eV, EI): m/z: 217 [M] , 202, 173, 131, 102, 98, 76.
2-Ethylhexyl p-methoxycarbonylcinnamate: 1H NMR (400 MHz, CDCl3,
258C): d ˆ 7.99 (d, 2J(H,H) ˆ 8.3 Hz, 2H), 7.64 (d, 2J(H,H) ˆ 16.1 Hz, 1H),
7.54 (d, 2J(H,H) ˆ 8.2 Hz, 2H), 6.48 (d, 2J(H,H) ˆ 16.1 Hz, 1H), 4.09 (m,
2
2H), 3.87 (s, 3H), 1.61 (quint, J(H,H) ˆ 6.0 Hz, 1H), 1.41 ± 1.26 (m, 8H),
0.87 (m, 6H); 13C{1H} NMR (101 MHz, CDCl3, 258C): d ˆ 166.6, 166.3,
142.9, 138.6, 131.2, 130.0, 127.8, 120.6, 67.0, 52.1, 38.7, 30.3, 28.9, 23.7, 22.9,
‡
14.0, 10.8; MS (70 eV, EI): m/z: 318 [M] , 287, 207, 189, 175, 145, 112, 70.
Acknowledgements
Experimental Section
The authors thank Mr. H. Schneider, Mr. N. Stoeckel (both TU München)
and Mrs. C. Fuhrmann (IfOK) for their excellent support for the
experimental work. Dr. M. Hateley (IfOK) is thanked for suggestions to
this manuscript. We thank Prof. K. Kühlein (former Hoechst AG), Dr. T.
Riermeier, Dr. H. Trauthwein (Aventis R&T), Dr. M. Eckert (Bayer AG),
Dr. J. Krauter (Degussa AG), Dr. A. Indolese, Dr. H.-U. Blaser (Solvias
AG) and Dr. F. Vollmüller (Clariant AG) for general discussions. A.Z.
General: All chemicals were commercially available and used without
further purification. DMAc was distilled over calcium hydride and stored
under argon. The coupling products synthesised were characterised using
GC/MS, 1H and 13C NMR.
General procedure: In an ACE pressure tube (Aldrich) aryl halide
(10 mmol), olefin (15 mmol), base (12 mmol), diethyleneglycol di-n-butyl
2914
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2914 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 13

Palladium Catalyst Systems
2908±2915
thanks the Studienstiftung des deutschen Volkes and the Max-Buchner-
Forschungsstiftung for grants. Generous support from the Deutsche
Forschungsgemeinschaft (DFG Be 1931/2-1 and 2-2) and gifts of precious
metals from Degussa AG are gratefully acknowledged.
Zhang, J. Huang, M. L. Trudell, S. P. Nolan, J. Org. Chem. 1999, 64,
3804; p) V. P. W. Böhm, C. W. K. Gstöttmayr, T. Weskamp, W. A.
Herrmann, J. Organomet. Chem. 2000, 595, 186; q) A. Zapf, M. Beller,
Chem. Eur. J. 2000, 6, 1830; r) A. Zapf, A. Ehrentraut, M. Beller,
Angew. Chem. 2000, 112, 4315, Angew. Chem. Int. Ed. 2000, 39, 4153.
[4] J. J. Bozell, C. E. Vogt, J. Am. Chem. Soc. 1988, 110, 2655.
[5] a) W. A. Herrmann, C. Broûmer, K. Öfele, C.-P. Reisinger, T.
Priermeier, M. Beller, H. Fischer, Angew. Chem. 1995, 107, 1989;
Angew. Chem. Int. Ed. Engl. 1995, 34, 1844; b) M. Beller, T. H.
Riermeier, Tetrahedron Lett. 1996, 37, 6535; c) W. A. Herrmann, C.
Broûmer, C.-P. Reisinger, T. H. Riermeier, K. Öfele, M. Beller, Chem.
Eur. J. 1997, 3, 1357; d) M. Beller, T. H. Riermeier, Eur. J. Inorg.
Chem. 1998, 29.
[6] Y. Ben-David, M. Portnoy, M. Gozin, D. Milstein, Organometallics
1992, 11, 1995.
[7] M. T. Reetz, G. Lohmer, R. Schwickardi, Angew. Chem. 1998, 110,
492; Angew. Chem. Int. Ed. 1998, 37, 481.
[8] A. F. Littke, G. C. Fu, J. Org. Chem. 1999, 64, 10.
[9] K. H. Shaughnessy, P. Kim, J. F. Hartwig, J. Am. Chem. Soc. 1999, 121,
2123.
[1] a) T. Mizoroki, K. Mori, A. Ozaki, Bull. Chem. Soc. Jpn. 1971, 44, 581;
b) R. F. Heck, J. P. Nolley, J. Org. Chem. 1972, 37, 2320; reviews: c) M.
Beller, T. H. Riermeier, G. Stark in Transition Metals for Organic
Synthesis, Vol. 1 (Eds.: M. Beller, C. Bolm), Wiley-VCH, Weinheim,
1998 p. 208; d) T. Jeffery, Adv. Met. Org. Chem. 1996, 153; e) A.
de Meijere, F. E. Meyer, Angew. Chem. 1994, 106, 2473; Angew.
Chem. Int. Ed. Engl. 1994, 33, 2379; f) S. Bräse, A. de Meijere in
Metal-Catalyzed Cross Coupling Reactions (Eds.: P. J. Stang, F.
Diederich), Wiley-VCH, Weinheim, 1988, p. 99; g) I. P. Beletskaya,
A. V. Cheprakov, Chem. Rev. 2000, 100, 3009.
[2] a) V. V. Grushin, H. Alper, Chem. Rev. 1994, 94, 1047; b) T. H.
Riermeier, A. Zapf, M. Beller, Top. Catal. 1997, 4, 301; c) R. Stürmer,
Angew. Chem. 1999, 111, 3509; Angew. Chem. Int. Ed. 1999, 38, 3307;
d) H. Gröger, J. Prakt. Chem. 2000, 342, 334.
[10] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165.
[3] Amination of aryl chlorides: a) J. F. Hartwig, M. Kawatsura, S. I.
Hauck, K. H. Shaughnessy, L. M. Alcazar-Roman, J. Org. Chem. 1999,
64, 5575; b) X. Bei, T. Uno, J. Norris, H. W. Turner, W. H. Weinberg,
A. S. Guram, Organometallics 1999, 18, 1840; c) X. Bei, A. S. Guram,
H. W. Turner, W. H. Weinberg, Tetrahedron Lett. 1999, 40, 1237; d) M.
Nishiyama, T. Yamamoto, Y. Koie, Tetrahedron Lett. 1998, 39, 617;
e) T. Yamamoto, M. Nishiyama, Y. Koie, Tetrahedron Lett. 1998, 39,
2367; f) M. Beller, T. H. Riermeier, C.-P. Reisinger, W. A. Herrmann,
Tetrahedron Lett. 1997, 38, 2073; Suzuki couplings of aryl chlorides:
g) M. Beller, H. Fischer, C. Broûmer, W. A. Herrmann, Angew. Chem.
1995, 107, 1992; Angew. Chem. Int. Ed. Engl. 1995, 34, 1848; h) A. F.
Indolese, Tetrahedron Lett. 1997, 38, 3513; i) S. Saito, S. Oh-tani, N.
Miyaura, J. Org. Chem. 1997, 62, 8024; j) A. F. Littke, C. Dai, G. C. Fu,
J. Am. Chem. Soc. 2000, 122, 4020; k) A. F. Littke, G. C. Fu, Angew.
Chem. 1998, 110, 3586; Angew. Chem. Int. Ed. 1998, 37, 3387; l) J. P.
Wolfe, S. L. Buchwald, Angew. Chem. 1999, 111, 2570; Angew. Chem.
Int. Ed. 1999, 38, 2413; m) J. P. Wolfe, R. A. Singer, B. H. Yang, S. L.
Buchwald, J. Am. Chem. Soc. 1999, 121, 9550; n) X. Bei, H. W. Turner,
W. H. Weinberg, A. S. Guram, J. Org. Chem. 1999, 64, 6797; o) C.
Â
[11] C. Amatore, E. Carre, A. Jutand, M. A. M'Barki, G. Meyer, Organo-
metallics 1995, 14, 5605.
[12] C. Amatore, A. Jutand, Acc. Chem. Res. 2000, 33, 314.
[13] a) C. Amatore, M. Azzabi, A. Jutand, J. Organomet. Chem. 1989, 363,
C41; b) T. Jeffery, Adv. Met. Org. Chem. 1996, 5, 154.
[14] R. J. Perry, B. D. Wilson, J. Org. Chem. 1996, 61, 7482.
[15] T. Jeffery, Tetrahedron 1996, 52, 10113.
[16] W. Cabri, I. Candiani, S. DeBernardinis, F. Francalanci, S. Penco, J.
Org. Chem. 1991, 56, 5796.
[17] C. A. Tolman, Chem. Rev. 1977, 77, 313.
[18] a) C. Amatore, M. Azzabi, A. Jutand, J. Am. Chem. Soc. 1991, 113,
8375; b) J.-F. Fauvarque, F. Pflüger, M. Troupel, J. Organomet. Chem.
1981, 208, 419.
[19] M. Beller, A. Zapf, Synlett 1998, 792.
[20] A. Ehrentraut, A. Zapf, M. Beller, Synlett 2000, 1589.
Received: December 14, 2000 [F2937]
Chem. Eur. J. 2001, 7, No. 13
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0713-2915 $ 17.50+.50/0
2915