NC-palladacycles as highly effective cheap precursors for the phosphine-free Heck reactions

Article abstract of DOI:10.1016/S0022-328X(00)00876-7

Eight cyclopalladated complexes of the formula [Pd₂(μ-L)₂(N-C)₂] (L=OAc, Cl; N-C=cyclometalled N donor: o-(2-pyridyl)phenyl, o-(2-pyridyloxy)phenyl, o-(2-pyridylmethyl)phenyl, o-(N,N-dimethylaminomethyl)phenyl, 8-quinolylmethyl and others) and a six-membered palladacycle with OC coordination (ligand related to 2-acetoamido-4-nitrophenyl), are highly efficient catalysts for the Heck arylation of olefins (styrene, ethyl acrylate) by aryl halides (iodobenzene, bromobenzene, 4-bromoacetophenone). These catalysts are air stable, easy to obtain from a vast number of readily available nitrogen containing molecules, are generally much cheaper than phosphine-ligated palladacycles, but as or more efficient than the latter. Turnover numbers (ton) of up to 4 100 000 and turnover frequencies (tof) up to 530 000 are achieved in the reaction of iodobenzene with ethyl acrylate. Bromobenzene undergoes the Heck reaction (ton=400-700; tof=5-30) in the presence of the promoter additive Bu₄NBr. The palladacycles are likely to operate in a common phosphine-free Pd(0)/Pd(II) catalytic cycle, while the differences between various types of palladacycle precursors are accounted for by the kinetics of the catalyst preactivation step.

Full text of DOI:10.1016/S0022-328X(00)00876-7

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 622 (2001) 89–96
NC-palladacycles as highly effective cheap precursors for the
phosphine-free Heck reactions
Irina P. Beletskaya *, Alexander N. Kashin, Natalia B. Karlstedt, Anton V. Mitin,
Andrei V. Cheprakov, Grigoriy M. Kazankov
Department of Chemistry, Moscow State Uni6ersity, 119899 Moscow, V-234, Russia
Received 3 August 2000; received in revised form 23 October 2000; accepted 23 October 2000
Abstract
Eight cyclopalladated complexes of the formula [Pd2(m-L)2(NꢀC)2] (L=OAc, Cl; NꢀC=cyclometalled N donor: o-(2-
pyridyl)phenyl, o-(2-pyridyloxy)phenyl, o-(2-pyridylmethyl)phenyl, o-(N,N-dimethylaminomethyl)phenyl, 8-quinolylmethyl and
others) and a six-membered palladacycle with OC coordination (ligand related to 2-acetoamido-4-nitrophenyl), are highly efficient
catalysts for the Heck arylation of olefins (styrene, ethyl acrylate) by aryl halides (iodobenzene, bromobenzene, 4-bromoacetophe-
none). These catalysts are air stable, easy to obtain from a vast number of readily available nitrogen containing molecules, are
generally much cheaper than phosphine-ligated palladacycles, but as or more efficient than the latter. Turnover numbers (ton) of
up to 4 100 000 and turnover frequencies (tof ) up to 530 000 are achieved in the reaction of iodobenzene with ethyl acrylate.
Bromobenzene undergoes the Heck reaction (ton=400–700; tof=5–30) in the presence of the promoter additive Bu4NBr. The
palladacycles are likely to operate in a common phosphine-free Pd(0)/Pd(II) catalytic cycle, while the differences between various
types of palladacycle precursors are accounted for by the kinetics of the catalyst preactivation step. © 2001 Elsevier Science B.V.
All rights reserved.
Keywords: Aryl halides; Catalysis; Heck reaction; Metallacycles; Palladium
SC-palladacycles are able to catalyze the reactions with
a wide selection of aryl halides, as well as vinyl halides
[6].
The palladacycles with «pincer» ligands of PCP and
SCS type (complexes 7–12, Scheme 2) turned out to be
also very efficient in the Heck reaction [7–11].
1. Introduction
In recent years the arylation of olefins with aryl
halides (Mizoroki–Heck reaction [1]) became not only
one of the most important ways to form Csp2ꢀC bond
2
sp
but also a benchmark to estimate the efficiency of a
catalytic system. Palladacycles with PC coordination
(complexes 1–3, Scheme 1) were found to be among the
most active precatalysts for the Heck reaction [2–4].
Complex 1 (Herrmann’s catalyst) is most extensively
studied and used in different types of Pd-catalyzed
reactions including the Heck reactions [2], the cross-
coupling reactions (Suzuki and Stille coupling) and the
arylation of amines. The use of nitrogen-based ligands
is very rare [5] though some of them were recently
demostrated to possess very high activity [5b,c]. The
* Corresponding author. Tel.: +7-095-9393618; fax: +7-095-
9393618.
E-mail address: beletska@org.chem.msu.ru (I.P. Beletskaya).
Scheme 1.
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00876-7

90
I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
2. Results and discussion
The comparison of catalytic activity of the palladacy-
cles was made in the reaction of iodobenzene with ethyl
acrylate in N,N-dimetylformamide (DMF) or dimethy-
lacetamide (DMAA) at 85°C (Table 1). This reaction is
largely selective and proceeds without formating
biphenyl. All mentioned complexes were found to be
able to catalyze the reaction to give high yields of ethyl
cinnamate (Scheme 4).
Under these conditions (0.1 mol% catalyst) the values
of ton are similar for the whole series of dimeric com-
plexes 13–20 to be practically equal to the value of ton
for catalyst 1 (entry 20, Table 1). However, the values
of turnover frequency (tof, mol product per mol palla-
dium h−1) vary strongly and show that complex 1 is
the most efficient one, while complexes 13, 15, 16, 20
follow it closely.
Scheme 2.
It is worth noting that for complex 15 tof value
determined at half life time of the reaction is even
higher than that of complex 1 (entries 5 and 20, Table
1).
A similar catalytic activity of the palladacycles has
been observed in the Heck reaction with styrene;
though, the values of tof are consistently lower (by
about three times) than for the reaction of ethyl acry-
late (entries 3, 9, 11, 13, 15, Table 1). The ratio of
diphenylethylene isomers determined by GLC is similar
to that observed for other catalyst systems [12]. The
conversion and sum of yields of the isomers agree
within experimental error limits (Scheme 5).
So far, we considered only the dimeric palladacycle
complexes. The monomeric complex 21 with pyridine
ligand showed similar catalytic activity as compared to
that of the respective dimeric complex 13 (entry 16 vs.
entry 1, Table 1). However, the presence of a Ph3P
ligand in complex 23 decreases the activity very
strongly (entry 19, Table 1; the reaction requires 22 h
for completion instead of 1.8 h for complex 1 or 3 h for
complex 13). The monomeric complex 22 with 2,2%-
dipyridyl ligand is inefficient in this reaction (entry 18,
Table 1). The addition of 2,2%-dipyridyl to the reaction
mixture in the presence of complex 21 also suppressed
the reaction (entry 17, Table 1).
Some of the palladacycles introduced above were
studied in the reaction with bromobenzenes. It was
shown that the reaction of 4-bromoacetophenone with
ethyl acrylate or styrene runs smoothly to give high
yields of the respective products though requiring a
prolonged heating at high temperature (140°C, Table 2)
(Scheme 6).
Scheme 3.
A characteristic feature of the palladacycles is the
thermostability, which makes it possible to perform the
reactions at temperatures above 100°C needed for less
reactive substrates (e.g. non-activated aryl bromides).
The highest turnover numbers (ton, mol product per
mol palladium) equal to 5.75.106 (180°C, 69 h, the yield
of 4-acetoxystilbene is 57.5%) [4] and 8.9.106 (180°C, 22
h, the yield of butyl cinnamate is 89%) [10] were
recorded so far for catalyst 3 in the reaction of 4-bro-
moacetophenone with styrene, and for catalyst 11 with
«pincer» PCP ligand in the reaction of iodobenzene
with butyl acrylate, respectively.
Here we report that many palladacycles with NC
coordination (complexes 13–19 and 21, Scheme 3), and
the complex with OC coordination 20, derived from
readily avalable and very inexpensive ligands can also
be used as catalyst precursors for the Heck reaction,
and some of them are very efficient allowing to carry
out the reactions under mild conditions [8].
Though we have not attempted to search for lowest
possible loads of the precatalyst in the reactions with
bromoarenes, the data already available show that the
activity of palladacycles under study is among the
highest reported so far. Thus, at a low load of pallada-

I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
91
Table 1
Heck arylation of olefins with iodobenzene in the presence of palladacycles a
c
d
g
Entry
Palladacycle
Olefin
t (h)
Conversion of PhI (%) b
ton
tof
tof1/2
1
2
3
4
13
13
13
14
Ethyl acrylate
Ethyl acrylate
Styrene
3.3
3
7.5
2.3
5.5
3
63 (58)
100 (93) e
100 (95) e
48
95 (89)
89 (82)
58
630
1000
1000
480
950
890
580
980
830
1000
720
1000
550
970
910
930
191
333
133
209
173
297
414
163
65
55
21
222
61
57
300
430 e
120 e
220
Ethyl acrylate
5
6
15
16
Ethyl acrylate
Ethyl acrylate
1000
400
1.4
6
98
83
7
8
9
17
17
17
18
18
19
19
20
20
21
21
22
23
1
Ethyl acrylate
Ethyl acrylate
Styrene
Ethyl acrylate
Styrene
Ethyl acrylate
Styrene
Ethyl acrylate
Styrene
Ethyl acrylate
Ethyl acrylate
Ethyl acrylate
Ethyl acrylate
Ethyl acrylate
12.8
18
35
4.5
9
17
30
2.7
15
6.5
2
86
100 (96) e
72 (69) e
100 e
55 e
10
11
12
13
14
15
16
17
18
19
20
330
67 e
53 e
97
91 e
30
340
66
93 e
700 e
42
120
100 (97)
87
1000
870
134
1 f
5
16
22
1.8
88
95 (90)
880
950
40
528
35
860
a Iodobenzene (0.1 mmol), olefin (0.12 mmol), tri-n-butylamine (0.1 mmol), catalyst (0.1 mol%), N,N-dimethylacetamide (20 ml), 85°C.
b Determined by titration of liberated iodide ions, see Section 4; data in parentheses refer to GLC yield of the olefinated product.
c ton-turnover number (mol product per mol catalyst).
d tof-turnover frequency (mol product per mol catalyst h−1).
e In N,N-dimethylformamide.
f In the presence of 2,2%-dipyridyl (0.1 mol% related to iodobenzene).
g tof in the range of 50% conversion of PhI.
Scheme 4.
ethyl acrylate with very low loads of complex 13 (Table
3).
cycle 13 at 0.002 mol% (entry 3, Table 2) the conversion
and yield of the cinnamate in the reaction of 4-bro-
moacetophenone with ethyl acrylate were high, and tof
reached 43500, a value comparable to that reported for
the reaction with complex 1 [2c].
The reaction of bromobenzene with ethyl acrylate or
styrene in the presence of the catalyst 13 is too slow
(entry 4, Table 2) but the addition of 20 mol% of
tetrabutylammonium bromide allowed one to achieve
72% and 38% conversion of bromobenzene after 24 h
or 72 h, respectively (entries 5 and 7, Table 2) (Scheme
7).
A huge ton value of about 1.105 can be achieved at
85°C with 0.001 mol% of the catalyst (entry 5, Table 3).
The highest values of ton (2–4).106 comparable with the
record figures obtained previously [4,10] were observed
for this reaction at 140°C in DMAA using (1–3).10−5
mol% of catalyst 13 (entries 7–9, Table 3).
In the case of the Heck reaction catalyzed by pallada-
cycles, a positive effect of tetrabutylammonium bro-
mide was observed by Herrmann et al. [2a,c]. The
efficiency of complex 13 under the conditions studied is
close to that of catalyst 1 in the reaction of bromoben-
zene with butyl acrylate (NaOAc, 140°C) [2c].
To test the higher limits of activity of the palladacy-
cles, we performed the reaction of iodobenzene with
Scheme 5.

92
I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
Table 2
Heck arylation of olefins with aryl bromides in the presence of palladacycles a
c
d
Entry
Pallada-cycle
13
mol% Olefin
Aryl bromide
t (h)
Conversion of PhI (%) b
ton
tof
1
2
3
4
5
6
7
8
9
0.05
0.01
0.002
0.1
0.1
0.1
Ethyl acrylate
4-Bromoacetophenone
4-Bromoacetophenone
4-Bromoacetophenone
Bromobenzene
Bromobenzene
4-Bromoacetophenone
Bromobenzene
24
24
24
24
24
64
72
24
64
100 (95)
98
87
2000
9800
43500
83
410
1810
Ethyl acrylate
Ethyl acrylate
Ethyl acrylate
Ethyl acrylate
Styrene
3
72 e
720
1000
380
9200
920
30
16
100 (98)
0.1
Styrene
38 e
5.3
16
0.01
0.1
Ethyl acrylate
Styrene
4-Bromoacetophenone
4-Bromoacetophenone
92
92 (86)
380
14
a Aryl bromide (0.25 mmol), olefin (0.3 mmol), tri-n-butylamine (0.25 mmol), N,N-dimethylacetamide (0.5 ml), 140°C.
b Determined by titration of liberated iodide ions, see Section 4; data in parentheses refer to GLC yield of the olefinated product.
c ton-turnover number (mol product per mol catalyst).
d tof-turnover frequency (mol product per mol catalyst h−1).
e In the presence of n-Bu4NBr.
Scheme 6.
Both ton and tof values (the letter reflects the relative
reaction rate) are equally important for a large-scale
implementation. The tof value for catalyst 13 is equal to
530000 (entry 7, Table 3) and it is a record figure
obtained so far for any palladium catalyst in the Heck
reaction. This puts the NC-palladacycles in one row
with the most active catalysts such as the Herrmann’s
catalyst.
It is interesting to consider how the values of ton and
tof change in the course of the reaction. The monitoring
of the reaction was carried out by potentiometric titra-
tion of the formed iodide ions. It was shown that the
concentration of iodide ions is in good agreement with
the product yield obtained by GLC technique. More-
over the titration method is more exact than GLC in
range of small conversion of PhI.The time dependence
of the ton and tof values for the most promising com-
plexes 13 and 15 in comparison with those of catalyst 1
is shown in Table 4 and Fig. 1.
In order to quantify the activity of a given catalytic
system, the values of tof are routinely used. In most
cases, these values are computed by dividing of the
cumulative ton value by the overall reaction time to
give a rough approximation of average activity. It
should be noted that the use of such averaged tof values
can be misleading, if two factors are not explicitly taken
into consideration: the induction period and flatness of
reaction rate profile at high conversions when the con-
centration of reagents becomes low. Both these regions
of a kinetic curve steal from the averaged tof values,
thus making the latter an inadequately underestimated
measure of the true catalytic activity. Therefore, cata-
lytic activity should be better quantified either by a
tangent of a kinetic curve at the point at which the
catalytic process starts after the induction period (the
initial tof ), or, if a smooth continuous kinetic curve is
not available, by tof averaged over shorter period of
time corresponding to incomplete conversion (e.g. 50%
conversion).
The inadequacy of tof values averaged over the
whole duration of reaction is clearly seen in the com-
parison of the behaviour of NC-palladacycles with Her-
rmann’s catalyst. As is seen from Fig. 1 complexes 1,
13, and 15 give closely matching averaged tof values,
while their actual behaviour is different. The reaction in
the presence of complex 1 starts immediately after
mixing of the reagents without any traceable induction
period, and runs to completion within ca. 2 h. The
reaction in the presence of complex 15 shows a similar
kinetics, though while the initial activity is even some-
what higher than the activity of complex 1, the conver-
Scheme 7.

I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
93
Table 3
Heck arylation of ethyl acrylate with iodobenzene with varying initial concentration of palladacycle 13 a
c
d
Entry
Solvent
DMF
T (°C)
mol% of 13
t (h)
Conversion of PhI (%) b
ton
tof
1
2
3
4
5
6
7
8
9
85
–
–
85
–
140
–
0.1
0.01
3
10
38
3.3
34
20
4
100 (93)
100
87
63 (58)
100
100
63
78 (72)
41
1000
10 000
87 000
330
1000
2300
190
0.001
0.1
0.001
0.0002
0.00003
–
DMAA
630
100 000
500 000
2 100 000
2 600 000
4 100 000
2900
25 000
530 000
350 000
170 000
–
–
7.5
24
0.00001
a Iodobenzene (0.1 mmol), ethyl acrylate (0.12 mmol), tri-n-butylamine (0.1 mmol), N,N-dimethylacetamide.
b Determined by titration of liberated iodide ions, see Section 4; data in parentheses refer to GLC yield of the olefinated product.
c ton-turnover number (mol product per mol catalyst).
d tof-turnover frequency (mol product per mol catalyst h−1).
Table 4
Dependence of iodobenzene conversion values and tof values on reaction time for Heck arylation of ethyl acrylate with iodobenzene in presence
of palladacycles a
c
Entry
1
Palladacycle
1
t (h)
Conversion of PhI (%) b
tof
tof1/2
860
0.25
0.5
1
1.5
2
24
43
77
90
96
99
960
860
770
600
480
330
3
2
3
15
13
0.25
0.5
1
1.5
2
33
51
69
76
80
88
1320
1020
690
510
400
1000
430
3
290
0.25
0.5
1
1.5
2
5
14
41
75
90
99
200
280
410
500
450
330
3
a Iodobenzene (0.1 mmol), ethyl acrylate (0.12 mmol), tri-n-butylamine (0.1 mmol), catalyst (0.1 mol%), N,N-dimethylacetamide (20 ml), 85°C.
b Determined by titration of liberated iodide ions.
c tof-turnover frequency (mol product per mol catalyst h−1).
sion remains incomplete, as the process is strongly
retarded after ca. 70% conversion. Thus, complex 15 gives
a more active, but simultaneously less stable catalytic
system, which is deactivated due to a fast depletion of
the catalytically active palladium species. On the other
hand, complex 13, which gives a catalytic system with an
averaged tof being identical to that observed for complex
1, while the kinetic curve displayed a relatively long
induction period. Nevertheless, after the induction period
it becomes obvious that the complex 13 gives a more
reactive catalytic system than complex 1, as both of the
processes run to completion within roughly the same time,
in spite of the time wasted for an induction period in the
case of complex 13. If the values of tof1/2 are used to
quantify the catalytic activity, this parameter is lower for
complex 13 than for complexes 1 or 15. So, we may
conclude that the differential initial tof values are better
for the description of the peak activity of the catalytic
system, while the use of tof1/2 permits to distinguish
between the processes with and without the induction
period.
The inductive period in the mentioned reaction can be
shortened or even completely eliminated by the addition
of an appropriate reductant. Fig. 2 shows the effect of
added NaBH4 on the kinetics of the reaction with complex
13. As we can see that in this case the curve almost loses
S-shape character and as a result the conversion of
iodobenzene grows by about 40% in the region of 120 min.

94
I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
Pd(0) in the absence of strong supporting ligands often
The induction period is often observed in the Heck
leads to immediate sedimentation of inactive palladium
black. The addition of reductant is therefore not recom-
mended for reactions with high initial loads of palla-
dium precatalyst (roughly more than 0.1 mol%). At
lower concentrations, at which the nucleation of palla-
dium clusters is slow due to lower probability of Pd(0)
encounters, the prereduction can indeed be helpful for
fast triggering of the catalytic process.
Both, the S-shaped kinetic curve and the effect of
reductant on shortening of the induction period unam-
biguosly show that the catalytic cycle in the reactions
under study is a conventional Pd(0)/Pd(II) Heck cycle
[1b,c]. Recently, several researchers put forward a hy-
pothesis that Heck reaction can be driven by Pd(II)/
Pd(IV) cycle [3,15], particularly in those cases when
palladium is bound to strong multidentate ligands,
which are believed not to be capable of full or partial
de-ligation, as, for instance, in the reactions catalyzed
by complexes with «pincer»-type ligands [7–11].
reaction in which palladium is introduced as a palladiu-
m(II) derivative. For example, there are short inductive
periods in the reaction of 4-bromoacetophenone with
butyl acrylate in the presence of the di-t-butylphos-
phino analogue of catalyst 1 [2c], and in the reaction of
4-chlorobromobenzene with a-methylstyrene in the
presence of catalyst 1 [13b]. It is eliminated by the
addition of appropriate reducing agents, cf., e.g. [14].
It should be noted that the prereduction must be
accomplished with proper care, as a fast reduction to
Still, so far there is no direct experimental evidence in
favour of a Pd(II)/Pd(IV) mechanism, while the evi-
dences against are plenty. All data usually cited in favor
of this mechanism, such as high thermal stability and
post-run recovery of the complexes, the rigidity of the
coordination spheres, etc, are not truly convincing. As
the reactions with precatalysts suspected for ability to
operate through a Pd(II)/Pd(IV) route are usually done
under very harsh conditions, the absolute stability of
even the most robust complexes is debatable. More-
over, the experiments with recovery are done with
higher initial loads of precatalyst than the experiments
intended to show a peak activity of a given system.
Since 100% recovery cannot be achieved, it is the part
of complex which escaped recovery that may indeed be
the source of actual catalyst, given the enormous cata-
lytic efficiency of such systems (e.g. for 1 mol% and
99% recovery we lose 0.01 mol% of Pd complex which
is well known to be more than enough for the reaction
to run at full pace). Thus, we believe that until a more
solid proof of a Pd(II)/Pd(IV) mechanism would be
available, the conventional Pd(0)/Pd(II) catalytic cycle
should be given a preference in the discussion of data
for all Heck reactions, except possibly those catalyzed
by truly robust «pincer»-type palladacycles with PCP-
coordination. All other types of palladacycles in the
preactivation stage preceding the entry into the Heck
cycle should undergo a full or partial disassembly with
a reduction of palladium to the zero-valent state.
The problem of how palladium(0) is generated in a
given catalytic system remains open. Some relevant
data have been so far obtained only for phosphine-as-
sisted catalytic processes [13,16]. In our case, the prere-
duction of Pd(II) with a simultaneous disassembly of
palladacycle can be effected by the olefin in a non-cata-
lytic Heck reaction [17] (Scheme 8).
Fig. 1. Comparison of different palladacycles in the Heck reaction:
conversion of iodobenzene against time. Reaction conditions:
iodobenzene (0.1 mmol), ethyl acrylate (0.12 mmol), tri-n-butylamine
(0.1 mmol), N,N-dimethylacetamide (20 ml); T=85°C; catalysts: 0.1
mmol (0.1 mol%) of the palladacycles 1, 13, and 15. The conversion is
determined by titration of liberated iodide ions (see Section 4).
Fig. 2. Conversion/time diagram for the Heck reaction of iodoben-
zene with ethyl acrylate catalyzed by the complex 9. Reaction condi-
tions: iodobenzene (0.1 mmol), ethyl acrylate (0.12 mmol),
tri-n-butylamine (0.1 mmol), N,N-dimethylacetamide (20 ml); T=
85°C; 0.1 mmol (0.1 mol%) of the palladacycle 13, and in the presence
of NaBH4 (0.2 mol%) (upper curve).
Scheme 8.

I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
95
Though Herrmann earlier disproved the involvement
of olefin insertion into the PdꢀC bond of palladacycles
with PC-coordination [2c], it can be relevant for NC-
palladacycles (by steric reasons the migratory insertion
of olefin is impossible without a prior dechelation of N
or P arm of the palladacycle, at least for palladacycles
with PdꢀCsp2 bonds). What indeed seems improbable
for a PC-system, may readily happen for the less
strongly bonded NC-system.
Bu4NBr. Huge ton values of up to 4 100 000 have been
achieved in the reaction of iodobenzene with ethyl
acrylate in the presence of catalyst 13. The tof value for
this catalytic system is equal to 530 000 h−1, which is a
record-making figure for any Heck reaction reported to
date.
The palladacycles with NC coordination showed a
comparable catalytic activity as the most famous Her-
rmann’s palladacycle.The data obtained can be ratio-
We may hypothesize that the principle mechanism of
the involvement of palladacycles in the Heck process is
a slow release of low-ligated highly active Pd(0) species.
The difference between various palladium precatalysts
is most probably accounted for by different rates of
Pd(0) release under the given conditions. Too fast a
release (which is the case when the precatalyst is a
relatively weak complex lacking strongly bonded lig-
ands capable of making Pd(0) state stable) is as ineffec-
tive as too slow a release (which is the case when the
precatalyst is a robust complex formed by strongly
bonded chelating ligands). The former leads to fast
accumulation of unstable Pd(0) species liable to aggre-
gation in the absence of excess of supporting phosphine
ligands or highly reactive substrates capable of scaveng-
ing them via the oxidative addition. Therefore, the first
type of precatalysts usually fail in the case of less
reactive bromoarenes. On the other hand, too robust
precursors (such as PC-palladacycles and particularly
«pincer»-type systems) require much higher tempera-
tures to launch the reaction and sustain a reasonable
rate.
Thus, the efficiency of a catalytic system utilizing
paladacycle precursors (as well as any other strongly
bonded complexes, e.g. those with carbene ligands [14])
depends on subtle differences in the kinetics of preacti-
vation. As such factors are practically unpredictable,
we believe that the search for highly effective catalytic
systems should be done by screening as many available
structures as possible. In this respect, NC-palladacycles
appear as particularly promising targets for research, as
such complexes are readily available from very simple
inexpensive nitrogen-containing molecules, the stock of
which is practically unlimited.
nalized within
a
conventional phosphine-free
Pd(0)/Pd(II) cycle. Though the exact mechanism for the
release of Pd(0) from the palladacycle carrier remains
unknown, the initial disassembly of the palladacycle
complex and liberation of highly reactive low ligated
Pd(0) species can be suggested to be the principal mode
of catalytic activity for the palladacycle and related
systems.
4. Experimental
4.1. General
Commercial chemicals and solvents were distilled or
recrystallized prior to use.
The conversion of aryl halide and the yields of
products were determined by the two independent
methods: (a) the potentiometric titration of halide ions
liberated during the reaction; (b) GLC analysis in the
significant cases. Quantitative GLC analysis was per-
formed on a Hewlett Packard gas chromatograph 5890
Series II Plus equipped with HP-1701 capillary column
(15 m×0.32 mm, 0.25 mm film) in conjunction with a
flame ionization detector (GC/FID); temperature: 35/
280°C, 7°C/min. The titration of halide ions with 0.03
M AgNO3/0.1 M HNO3 was carried out on a Radiome-
ter Titrator TTT1c, Titrigraph SBR2c, autoburet
ABU1 (0.25 ml) using an indicator electrode Selectrode
F1212S.The accuracy of this method is within 2%
range. The validity of the potentiometric titration
method has been established by blind runs carried out
under conditions identical to the conditions of the Heck
reaction but in the absence of olefin. The formation of
iodide ions in the blind runs does not exceed the
instrumental error. The yields determined by GLC and
potentiometric titration of halide ions were shown to
agree within 7% range.
3. Conclusion
Ten palladacycles with NC coordination and one
with OC coordination have been studied as precursors
of catalysts in the Heck arylation of olefins with aryl
halides. Nine of them show an excellent efficiency in the
reactions of iodobenzene with styrene and ethyl acry-
late. The most active catalysts 13 and 16 were used also
in the reaction of styrene or ethyl acrylate with 4-bro-
moacetophenone and bromobenzene. The reaction with
less reactive bromobenzene require the addition of
The reactions were carried out either under positive
argon pressure in the reactor equipped with a mercury
valve, or in a sealed and evacuated (10−4 mmHg) glass
ampoule for slower reactions. N,N-dimethylacetamide
and N,N-dimetylformamide were carefully dried ac-
cording to known procedures. The solutions of cata-
lysts were prepared for each reaction and used fresh
only once.

96
I.P. Beletskaya et al. / Journal of Organometallic Chemistry 622 (2001) 89–96
4.2. Preparation of palladacycles
References
[1] (a) T. Mizoroki, K. Mori, A. Ozaki, Bull. Chem. Soc. Jpn. 44
(1971) 581. (b) R.F.Heck, J.P. Jr. Nolley, J. Org. Chem. 37
(1972) 2320. (c) R.F. Heck, Acc. Chem. Res. 12 (1979) 146. (d)
W. Cabri, I. Candiani, Acc. Chem. Res. 28 (1995) 2.
[2] (a) W.A. Herrmann, V.P.W. Bohm, C.-P. Reisinger, J.
Organomet. Chem. 576 (1999) 23. (b) W.A. Herrmann, C.
Brossmer, K. Ofele, C.-P. Reisinger, T. Priermeier, M. Beller, H.
Fischer, Angew. Chem. Int. Ed. Engl. 34 (1995) 1844. (c) W.A.
Herrmann, C. Brossmer, C.-P. Reisinger, T.H. Riermeier, K.
Ofele, M. Beller, Chem. Eur. J. 3 (1997) 1357. (d) M. Beller,
T.H. Riermeier, S. Haber, H.-J. Kleiner, W.A. Herrmann Chem.
Ber. 129 (1996) 1259.
Literature procedures were used for synthesis of com-
plexes 1 [2c], 13 [18], 14 [19], 15 [20], 16 [21], 17 [22], 18
[23], 19 [24], 20 [25], 21 [26], 23 [18].
4.3. [8-Quinolylmethyl](2,2%-dipyridyl)palladium(II)
chloride (22)
Complex 22 was prepared according to a previously
described procedure [24b]; Anal. Found: C, 54.21; H,
3.72; N, 9.78. Calc for C20H16N3PdCl (440.24): C,
54.57; H, 3.66; N, 9.54%.
[3] B.L. Shaw, S.D. Perera, E.A. Staley, Chem. Commun. (1998)
1361.
[4] D.A. Albisson, R.B. Bedford, P.N. Scully, Tetrahedron Lett. 39
(1998) 9793.
[5] (a) W. Cabri, I. Candiani, A. Bedeschi, R. Santi, Synlett. (1992)
871. (b) M. Ohff, A. Ohff, D. Milstein Chem. Commun. (1999)
357. (c) D.A. Alonso, C. Najera, M.C. Pacheco, Org. Lett. 2
(2000) 1823.
[6] A.S. Gruber, D. Zim, G. Eberling, A.L. Monteiro, J. Dupont,
Org. Lett. 2 (2000) 1287.
[7] M. Ohff, A. Ohff, M.E. van der Boom, D. Milstein, J. Am.
Chem. Soc. 119 (1997) 11687.
[8] I.P. Beletskaya, A.N. Kashin, N.B. Karlstedt, A.V. Chuchuri-
jkin, 10th IUPAC Symposium on Organo-Metallic Chemistry
directed toward Organic Synthesis, July 18–22, 1999, Versailles,
France, SL-28.
[9] K. Kiewel, Y. Liu, D.E. Bergbreiter, G.A. Sulikowski, Tetrahe-
dron Lett. 40 (1999) 8945.
[10] F. Miyaazaki, K. Yamaguchi, M. Shibasaki, Tetrahedron Lett.
40 (1999) 7379.
[11] D.E. Bergbreiter, P.L. Osburn, Y.-S. Liu, J. Am. Chem. Soc. 121
(1999) 9531.
[12] (a) M.T. Reetz, G. Lohmer, R. Schwickardi, Angew. Chem. Int.
Ed. 37 (1998) 481. (b) M.T. Reetz, E. Westermann, R. Lohmer,
G. Lohmer, Tetrahedron Lett. 39 (1998) 8449.
[13] (a) J. Louie, J.F. Hartwig, Angew. Chem. Int. Ed. Engl. 35
(1996) 2359. (b) M. Beller, T.H. Riermeier, Eur. J. Inorg. Chem.
(1998) 29.
[14] W.A. Herrmann, M. Elison, J. Fischer, C. Kocher, G.R.J. Artus,
Angew. Chem. Int. Ed. Engl. 34 (1995) 2371.
[15] B.L. Shaw, New J. Chem. 22 (1998) 77.
[16] F.-T. Luo, A. Jeevanandam, M.K. Basu, Tetrahedron Lett. 39
(1998) 7939.
[17] J. Tsuji, Acc. Chem. Res. 2 (1969) 144.
[18] G.E. Hartwell, R.V. Lawrence, M.J. Smas, J. Chem. Soc. (D).
(1970) 912.
[19] A. Kasahara, Bull. Chem. Soc. Jpn. 41 (1968) 1272.
[20] K. Hiraki, Y. Fuchita, K. Takechi, Inorg. Chem. 20 (1981) 4316.
[21] D.J. de Geest, B.J. O’Keefe, P.J. Steel, J. Organomet. Chem. 579
(1999) 97.
4.4. General procedure for the Heck reaction
Catalyst solution: the appropriate palladacycle (see
Scheme 3) was dissolved in 1 ml N,N-dimethylac-
etamide. The solution was purged with argon prior to
use. Reaction mixture: In a thermostated three-necked
flask installed on a magnetic stirrer, aryl halide (0.1
mmol), olefin (0.12 mmol), tri-n-butylamine (240 ml, 1
mmol), the solution of 1,4-di-t-butylbenzene (GLC in-
ternal standard) and N,N-dimethylacetamide (20 ml)
were placed while the flask being continously flushed
with argon. The reaction mixture was vigorously stirred
for 5 min and heated to 85°C. The catalyst solution
(about 10 ml) was injected by syringe into the preheated
reaction mixture. For kinetic studies, samples of 20–50
ml were withdrawn from the reaction mixture, mixed
with 1.5 ml 0.1 M HNO3 and titrated. To determine the
product yields aliquots of the reaction mixture were
diluted with 5% hydrochloric acid (1:5) and extracted
with methylene chloride (3×2 ml). The organic phases
were dried with anhydrous magnesium sulfate and ana-
lyzed by GLC.
The reactions with minimal loads of the palladium
catalyst were carried out in sealed glass ampoules. Aryl
halide (0.25 mmol), olefin (0.3 mmol), tri-n-butylamine
(0.25 mmol), a solution of 1,4-di-t-butylbenzene (GC
standard), N,N-dimethylacetamide (0.5 ml) and the cat-
alyst solution (about 10 ml) were placed in an ampoule.
The reaction mixture was thoroughly degassed with two
freeze-pump-thaw cycles, sealed and placed in a heating
bath. The work-up and analysis were performed as
specified above.
[22] A.G. Cope, E.C. Friedrich, C. Edwin, J. Am. Chem. Soc. 90
(1968) 909.
[23] A.D. Ryabov, G.M. Kazankov, Koord. Khim. (in russian) 12
(1986) 540.
[24] (a) H. Onoue, K. Minami, K. Nakagawa, Bull. Chem. Soc. Jpn.
43 (1970) 3480. (b) A.D. Ryabov, G.M. Kazankov, A.K. Yat-
simirsky, L.G. Kuz’mina, O. Yu. Burtseva, N.V. Dvortsova,
V.A. Polyakov, Inorg. Chem. 31 (1992) 3083.
[25] H. Horino, N. Inone, J. Org. Chem. 46 (1981) 4416.
[26] G.M. Kazankov, P.A. Posel’enov, A.D. Ryabov, A.K. Yat-
simirsky, Metallorg. Khim. (in russian) 4 (1991) 45.
Acknowledgements
Financial support by the Russian Foundation for
Basic Research (project Nos. 00-03-32766, 00-15-97406
and AO115) is gratefully acknowledged.
.