Heck reaction catalyzed by phospha-palladacycles in non-aqueous ionic liquids

Article abstract of DOI:10.1016/S0022-328X(98)00941-3

Phospha-palladacycles are among the most powerful palladium catalyst systems for the Heck reaction. We have shown the use of non-aqueous ionic liquids (NAILs) as an alternative to traditional molecular solvents for this reaction, with the phospha-palladacycle catalysts resulting in easy product separation, possible catalyst recycling and further increases in catalyst productivity. Preliminary results obtained with bromo- and chloro arenes are presented.

Full text of DOI:10.1016/S0022-328X(98)00941-3

Journal of Organometallic Chemistry 572 (1999) 141–145
Preliminary communication
Heck reaction catalyzed by phospha-palladacycles in non-aqueous ionic
liquidsꢀ
Wolfgang A. Herrmann *, Volker P.W. Bo¨hm
Anorganisch-chemisches Institut der Technischen Uni6ersita¨t Mu¨nchen, Lichtenbergstrasse 4, D-85747 Garching bei Mu¨nchen, Germany
Received 25 August 1998
Abstract
Phospha-palladacycles are among the most powerful palladium catalyst systems for the Heck reaction. We have shown the use
of non-aqueous ionic liquids (NAILs) as an alternative to traditional molecular solvents for this reaction, with the phospha-pal-
ladacycle catalysts resulting in easy product separation, possible catalyst recycling and further increases in catalyst productivity.
Preliminary results obtained with bromo- and chloro arenes are presented. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Palladium; Palladacycles; Catalysis; Heck reaction; Non-aqueous ionic liquids
1. Introduction
We have introduced the cyclometallated complex
trans-di(v-acetato)-bis[o-(di-o-tolylphosphino)benzyl]
dipalladium(II) 1 as an efficient catalyst for the Heck
reaction with aryl bromides and electron-deficient aryl
chlorides [4]. Although it is possible to retain complex 1
unchanged in yields up to 90 % after catalysis, this has
never been a useful protocol for catalyst recycling be-
cause of the need for recrystallization and column
chromatography.
Organopalladium catalyzed carbon–carbon bond
forming reactions are versatile tools in synthetic chem-
istry [1]. In particular the vinylation of aryl halides—
commonly called the Heck reaction—has received
increased attention in the literature because of its enor-
mous synthetic potential for generating carbon–carbon
bonds and its tolerance towards a wide range of func-
tional groups [2,3]. Attempts to achieve technical appli-
cation of this reaction have failed in most cases due to
the necessity of expensive aryl iodides or of high
amounts of palladium catalyst. The former can be
overcome by the use of more active catalysts being able
to activate the more economic aryl bromides and chlo-
rides whereas the latter can be overcome by recycling
the catalyst.
ꢀ Coordination chemistry and mechanism of metal-catalyzed C–C
coupling reactions, Part 11; for Part 10 see R. Schmid, W.A. Herr-
mann, G. Frenking, Organometallics 16 (1997) 701.
* Corresponding author. Tel.: +49-89-289-13080; Fax: +49-89-
289-13473; e-mail: lit@arthur.anorg.chemie.tu-muenchen.de.
As the Heck reaction is normally performed in polar
solvents like acetonitrile, DMF, DMAc or NMP [3]
and since it was known that salt additives can activate
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)00941-3

142
W.A. Herrmann, V.P.W. Bo¨hm / Journal of Organometallic Chemistry 572 (1999) 141–145
Table 1
Heck reaction of chlorobenzene with styrene and 0.5 mol% palladacycle 1 as catalysta
Solvent (NAIL)
Base
Additive (mol%)
t (h)
Conversion (%)b
Yield (%)c
Isomer ratio (1,1:Z:E)
1d
2
3
4
5
6
7
8
9
10
11
12
13
[NBu4]Br
Na[OAc]
Na[OAc]
–
Na[OAc]
Na[OAc]
Na[OAc]
Na[OAc]
Et3N
K2CO3
K3PO4
Na[OAc]
Et3N
–
16
18
14
19
16
16
20
16
16
16
17
16
24
28
52
54
22
20
24
2
46
42
40
44
44
43
26 (93)
51 (98)
54 (99)
22 (99)
19 (95)
21 (88)
2 (99)
46 (99)
41 (98)
38 (95)
36 (82)
32 (73)
43 (99)
7:2:91
4:0:96
6:2:92
8:10:82
9:2:89
10:3:87
1:0:99
15:0:85
3:1:96
3:0:97
10:2:88
11:1:88
6:2:92
[NBu4]Br
[NBu4]OAc
MPIMBre
TBHDPe
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
[PPh4]Cl (6)
–
TPMPCe
f
[NBu4]Br
[NBu4]Br
[NBu4]Br
[NBu4]Br
[NBu4]Br
[NBu4]Br
–
Na[OAc]
[PPh4]Cl (6)
DMGe (12)
[PPh4]Cl (20)
[AsPh4]Cl (6)
[AsPh4]Cl (20)
14
15
16
[NBu4]Br
[NBu4]Br
[NBu4]Br
Na[OAc]
Na[OAc]
Na[OAc]
24
17
24
54
62
78
53 (98)
62 (99)
77 (99)
11:1:88
4:1:95
6:2:92
a One equivalent PhCl, 1.5 equivalents styrene, 1.2 equivalents base, 1 mol% Pd (0.5 mol% 1); T=150°C.
b GC conversion of chlorobenzene using diethyleneglycol-di-n-butylether as internal standard.
c GC yield based on PhCl; yield based on conversion of PhCl in brackets.
d 1 mol% Pd(OAc)2 as catalyst.
e MPIMBr=1-methyl-3-propylimidazolium bromide; DMG=N,N-dimethylglycine; TBHDP=tri-n-butyl-n-hexadecylphosphonium bromide;
TPMPC=triphenylmethylphosphonium chloride; TPMPB=triphenylmethylphosphonium bromide.
f TPMPC/TPMPB (2:1) was used as solvent mixture.
hexadecylphosphonium bromide was described as a
and stabilize the catalytically active palladium species
[5], the use of non-aqueous ionic liquids (NAILs), i.e.
molten salts, seemed to be a promising alternative to
conventional molecular solvents [6]. The use of ionic
liquids would have an advantage over conventional
reaction conditions if the product could be distilled
from the ‘solvent’ and the still active, stable catalyst
thus giving the opportunity to recycle the whole cata-
lyst–solvent system. This concept has already been
tested for bromobenzene and butyl acrylate, but only
with two different salts and catalysts [7]. Also, no
details were given on the possibility of recycling the
catalyst system in this report.
superior reaction medium with bis(triphenylphos-
phine)palladium(II) dichloride as catalyst [7], but re-
sulted in lower turnover numbers (TONs) with catalyst
1. Sodium acetate, triethylamine, potassium carbonate
and potassium phosphate were tested as base. Na[OAc]
performed best, without exhibiting large differences in
product yields compared to NEt3 (Table 1, entries 2
and 8). Comparison of palladacycle 1 with palladiu-
m(II) acetate showed that it is necessary to use a
phosphine-stabilized catalyst rather than a palladium
salt (Table 1, entry 1). Not only a lower TON but also
the formation of palladium black which could hamper
recycling of the catalyst makes Pd(OAc)2 no alternative
to palladacycle 1.
2. Results and discussion
Using catalyst 1 does not lead to visible palladium
black formation in most of the described media even in
the reaction of chlorobenzene with styrene at 150°C.
The addition of tetraphenylphosphonium chloride,
[PPh4]Cl, which had previously shown to be a superior
additive for the Heck reaction [9], caused an increase in
TON with chlorobenzene. The analogous arsonium
salt, [AsPh4]Cl, proved to be an even better additive in
The efficiency of various molten salts as solvent for
the Heck reaction was tested by coupling chlorobenzene
and styrene with 0.5 mol% palladacycle 1 (1 mol% Pd)
[8]. In terms of productivity and stability of the cata-
lyst, tetra-n-butylammonium bromide performed best
as compared to other tested salts (Table 1). Tributyl-

W.A. Herrmann, V.P.W. Bo¨hm / Journal of Organometallic Chemistry 572 (1999) 141–145
143
Table 2
Heck reaction of aryl halides with olefins and palladacycle 1 as catalyst in [NBu4]Bra
b
R
X
Pd (mol%) Additive (mol%) T (°C)
t (h)
Conversionc (%)
Yield/isomer ratio
(%)/(1,1:Z:E)d
TON (mol
product·mol Pd−1
)
4-COCH3
4-H
4-H
4-n-Bu
4-OCH3
2-thiophenyl
Br 0.0001
Br
Br 0.1
Br 0.1
–
–
–
–
–
120
130
120
130
120
60
12
16
15
18
16
\99
\99
\99
99
\99
93
\99 (7:1:92)
\99 (6:1:93)
99 (5:1:94)
99 (15:1:84)
99 (1:0:99)
92 (9:5:86)
1 000 000
100
990
990
50
1
Br
2
Br 0.1
[AsPh4]Cl (0.1) 130
[PPh4]Cl (1) 150
[AsPh4]Cl (0.1) 150
[PPh4]Cl (1)
[AsPh4]Cl (20)
[PPh4]Cl (1)
[PPh4]C1 (2)
[AsPh4]Cl (2)
920
4-NO2
4-COCH3
4-CF3
Cl
1
16
15
15
16
66
14
16
\99
99
99
97
39
\99 (3:5:92)
99 (8:2:90)
98 (5:2:93)
96 (4:1:95)
39 (3:1:96)
18 (22:2:76)
51 (1:4:95)
100
990
98
48
390
5
Cl 0.1
Cl
Cl
1
2
150
150
150
150
160
4-H
4-He
Cl 0.1
4-OCH3
2-thiophenyl
Cl
Cl
4
1
20
58
51
a One equivalent ArX, 1.5 equivalents olefin, 1.2 equivalents Na[OAc], catalyst; [NBu4]Br.
b Catalyst amounts given refer to Pd; for the amount of 1, the value given has to be halved.
c GC conversion of aryl halide using diethyleneglycol-di-n-butylether as internal standard.
d GC yield based on ArX.
e With 2-ethylhexyl acrylate as the olefin.
molten [NBu4]Br although a darkening of the solution
was observed in contrast to the addition of [PPh4]Cl.
Thus, chlorobenzene could be coupled with styrene in
almost quantitative yield by 2 mol% palladium (Table
2). These observations indicate that [PPh4]Cl constitutes
a stabilizing and activating additive whereas [AsPh4]Cl
activates the catalyst but does not stabilize it. The dark
solution obtained could also indicate the formation of
palladium(0) aggregates implying that the reaction took
place at the surface of palladium clusters, a hypothesis
which is currently under investigation. Nevertheless un-
til now, colloidal palladium was not capable of effi-
ciently coupling chlorobenzene [10], thus suggesting
that at least with PhCl the reaction is not mediated by
palladium clusters. Using [PPh4]Cl and [AsPh4]Cl addi-
tives in the Heck reaction increases selectivity towards
vinylation, i.e. less dehalogenation and biaryl formation
occurs. Unfortunately, at the same time the reaction
suffers from the formation of scrambled by-products in
B1–15% yield with activated aryl chlorides and from
20 up to 50% yield in the case of deactivated aryl
chlorides depending on the substrates, the additive con-
centration and the reaction conditions. The isomer
distributions of 1,1-diphenylethylene, (Z)-stilbene and
(E)-stilbene are in the expected ratios with the (E)-iso-
mer always being formed as the main product. The
addition of N,N-dimethylglycine (DMG) had no fur-
ther influence on the isomer distribution in contrast to
literature reports [9].
The reaction of aryl bromides and aryl chlorides with
styrene was also tested under the best conditions found
for the reaction of chlorobenzene (Table 2). Activated
aryl bromides like 4-bromoacetophenone react at low
catalyst concentrations quantitatively. Thus, TONs of
1 000 000 [(mol product)·(mol palladium)−1] have been
achieved for the reaction of 4-bromoacetophenone with
styrene. Deactivated aryl bromides like 4-bromoanisole
require higher catalyst concentrations for quantitative
yields. Activated aryl chlorides could be coupled quan-
titatively with good TONs of up to 1000. In these cases
the use of NAILs proved to be much more effective
than molecular solvents [4]. Nevertheless, deactivated
aryl chlorides like 4-chloroanisole still leave room for
further improvement.
The stability of catalyst 1 in molten [NBu4]Br sug-
gested a possible recycling procedure [11]. The catalytic
activity of complex 1 after recycling was tested by
reacting bromobenzene with styrene and 1 mol% of
palladium in [NBu4]Br. Recycling of the solvent and the
catalyst 1 was achieved by distillation of the reactants
and products from solvent and catalyst in vacuo. Only
NaBr as a stoichiometric salt by-product was left in the
solvent–catalyst system. After the fourth run visible
palladium black formation occurred and after the sixth

144
W.A. Herrmann, V.P.W. Bo¨hm / Journal of Organometallic Chemistry 572 (1999) 141–145
Fig. 1. Heck reaction of bromobenzene with styrene and 0.5 mol% palladacycle 1 as catalyst (above). Recycling of solvent and catalyst (below).
Conditions: 1 equivalent PhBr (382 mg), 1.5 equivalents styrene (380 mg), 1.2 equivalents Na[OAc] (239 mg), 0.5 mol% 1; 3.0 g [NBu4]Br;
T=130°C; t=12 h. Isolated yields of analytically pure stilbene by distillation.
at the same time. Furthermore, NAIL protocols re-
quire less solvent due to the good solubility of
Na[OAc] base and salt additives. Thus, the price for
the solvent per reaction is comparable to NMP or
DMAc.
Comparison of the yields obtained with palladacycle
1 in [NBu4]Br with those obtained in DMAc shows
that aryl bromides can be coupled with the same high
TONs. In the case of aryl chlorides even an enhance-
ment in activity was observed most probably due to
increased thermal stability of the catalytically active
species at elevated temperatures of 150°C and higher.
Further investigations concerning comparisons with
other established palladium catalysts, activation of aryl
chlorides and the detection of the catalytically active
species are currently in progress.
run the melt became more viscous but catalysis could
still be performed with excellent yields in further runs.
Decomposition of the catalyst and the formation of
palladium black could have been caused by either a
long term, intrinsic catalyst instability or by a NaBr
by-product promoted catalyst destabilisation. Further
investigations are in progress. A graphical comparison
of the catalytic activity of complex 1 over consecutive
runs is shown in Fig. 1.
3. Conclusions
Molten salts as media for the Heck reaction have
various advantages. As certain salt additives stabilize
and activate the catalyst, the use of these additives as
‘solvents’ increases the stabilization and activation to a
further extent. Inhibition of catalysis by the blocking
of essential coordination sites on the palladium center
has not been observed. Although NAIL protocols are
homogeneous one-phase reaction conditions, catalyst
recycling is possible like in two-phase- or in heteroge-
neous catalysis. Thus, the advantages of homogeneous
catalysis like high selectivity and activity of the cata-
lysts as well as the possibility of recycling can be used
Acknowledgements
This research was supported by the German Science
Foundation (DFG), Degussa AG (palladium salt
grants) and the Fonds der Chemischen Industrie (stu-
dentship for VPWB). We would like to thank Dr
M.G. Gardiner for stimulating and helpful discussion.

W.A. Herrmann, V.P.W. Bo¨hm / Journal of Organometallic Chemistry 572 (1999) 141–145
145
[6] Y. Chauvin, L. Mussmann, H. Olivier, Angew. Chem. 107 (1995)
2941; Angew. Chem. Int. Ed. Engl. 34 (1995) 2698.
References
[7] D.E. Kaufmann, M. Nouroozian, H. Henze, Synlett 11 (1996)
1091.
[1] (a) K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, 3rd
ed., VCH, Weinheim, 1988. (b) K.C. Nicolaou, E.J. Sorensen,
Classics in Total Synthesis, VCH, Weinheim, 1996.
[2] (a) R.F. Heck, J.P. Nolley, J. Org. Chem. 37 (1972) 2320. (b) T.
Mizoroki, K. Mori, A. Ozaki, Bull. Chem. Soc. Jpn. 44 (1971)
581.
[3] Reviews: (a) R.F. Heck, in: B.M. Trost, I. Fleming (Eds.),
Comprehensive Organic Synthesis, vol. 4, Pergamon, Oxford,
1991, pp. 833 ff. (b) A. de Meijere, Angew. Chem. Int. Ed. Engl.
33 (1994) 2379. (c) W.A. Herrmann, in: B. Cornils, W.A. Herr-
mann (Eds.), Applied Homogeneous Catalysis with
Organometallic Compounds, VCH, Weinheim, 1996, pp. 712 ff.
[8] General experimental procedure: 1.5 g of salt are stirred in vacuo
at 130°C or 10 K over its melting point for 1 h in order to
remove water and oxygen. After cooling to room temperature
1.0 mmol aryl halide, 1.5 mmol olefin, 1.2 mmol base, 0.05 g
diethyleneglycol-di-n-butylether (GC standard) and catalyst 1
are added (all have been previously degassed) under an inert
reaction atmosphere. After completion of the reaction the mix-
ture was diluted with water, the products extracted with CH2Cl2
and filtered. Yields were determined by GC/MS.
[9] M.T. Reetz, G. Lohmer, R. Schwickardi, Angew. Chem. 110
(1998) 492; Angew. Chem. Int. Ed. Engl. 37 (1998) 481.
[10] (a) M. Beller, H. Fischer, K. Ku¨hlein, C.-P. Reisinger, W.A.
Herrmann, J. Organomet. Chem. 520 (1996) 257. (b) M.T. Reetz,
R. Breinbauer, K. Wanninger, Chem. Commun. (1996) 1921.
[11] For the experiments concerning recycling, the reactants and
products were distilled from the catalyst and the solvent at
130°C/10−2 mbar. Yields given are of the isolated, analytically
pure product isomers.
8
[4] (a) W.A. Herrmann, C. Broßmer, K. Ofele, C.-P. Reisinger, T.
Priermeier, M. Beller, H. Fischer, Angew. Chemie 107 (1995)
1989; Angew. Chem. Int. Ed. Engl. 34 (1995) 1844. (b) W.A.
Herrmann, C. Broßmer, C.-P. Reisinger, T.H. Riermeier, K.
8
Ofele, M. Beller, J. Eur. Chem. 3 (1997) 1357.
[5] (a) T. Jeffery, Tetrahedron Lett. 26 (1985) 2667. (b) T. Jeffrey, J.
Chem. Soc. Chem. Commun. (1984) 1287.
.