Homeopathic ligand-free palladium as a catalyst in the heck reaction. A comparison with a palladacycle

Article abstract of DOI:10.1021/ol035184b

(Matrix presented) Ligand-free Pd(OAc)₂ can be used as a catalyst in the Heck reaction of aryl bromides as long as the amount of catalyst is kept between 0.01 and 0.1 mol %. At higher concentrations palladium black forms and the reaction stops. The actual catalyst is monomeric. Palladacycles merely serve as a source of ligand-free palladium in Heck reactions of aryl bromides.

Full text of DOI:10.1021/ol035184b

ORGANIC
LETTERS
2003
Vol. 5, No. 18
3285-3288
Homeopathic Ligand-Free Palladium as
a Catalyst in the Heck Reaction. A
Comparison with a Palladacycle
Andre´ H. M. de Vries,* Jan M. C. A. Mulders, John H. M. Mommers,
Huub J. W. Henderickx, and Johannes G. de Vries*
DSM Pharma Chemicals-AdVanced Synthesis, Catalysis & DeVelopment, P.O. Box 18,
6160 MD Geleen, The Netherlands
andre.Vries-de@dsm.com
Received June 26, 2003
ABSTRACT
Ligand-free Pd(OAc)2 can be used as a catalyst in the Heck reaction of aryl bromides as long as the amount of catalyst is kept between 0.01
and 0.1 mol %. At higher concentrations palladium black forms and the reaction stops. The actual catalyst is monomeric. Palladacycles merely
serve as a source of ligand-free palladium in Heck reactions of aryl bromides.
The Heck reaction has recently been the focus of attention
in both the academic1 and industrial2 community. This has
been fuelled by the discovery of new generations of catalysts,
such as palladacycles and pincers, and more recently the
bulky electron-rich ligands for coupling reactions on aryl
chlorides.3
Our research has focused on the reduction of costs of these
reactions by the creation of highly active catalysts,4 and on
the reduction of waste by developing a halide-free Heck
reaction based on the use of aromatic anhydrides as arylating
agents.5 From the perspective of process development, there
is a strong drive to simplify these reactions as much as
possible. In this respect, the presence of phosphine ligands
can be a nuisance, particularly at high catalyst loadings, as
they are often hard to separate from the product.
Mainly due to the work of Jeffery,6 it is known that ligand-
free palladium7 can be used for the Heck reaction on aryl
iodides. It is also possible to apply ligand-free palladium
for Heck reactions in water,8 or on less usual leaving groups
such as diazonium salts, aroyl halides, anhydrides, and other
(1) (a) Heck, R. F. Org. React. 1982, 27, 345. (b) Heck, R. F. In
ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon
Press: Oxford, UK, 1991; Vol 4, Chapter 4.3, p 833. (c) Cabri, W.;
Candiani, I. Acc. Chem. Res. 1995, 28, 2. (d) Bra¨se, S.; de Meijere, A. In
Metal-Catalyzed Cross-Coupling Reactions; Diederich, F., Stang, P. J., Eds.;
Wiley-VCH: Weinheim, Germany, 1998; p 99. (e) Beller, M.; Riermeier,
T. H.; Stark, G. In Transition Metals for Organic Synthesis; Building Blocks
and Fine Chemicals; Beller, M., Bolm, C., Eds.; Wiley-VCH: Weinheim,
Germany, 1998; Vol 1, p 208. (f) Beletskaya, I. P.; Cheprakov, A. V. Chem.
ReV. 2000, 100, 3009. (g) Whitcombe, N. J.; Hii, K. K. (M.); Gibson, S. E.
Tetrahedron 2001, 57, 7449.
(5) (a) Stephan, M. S.; Teunissen, A. J. J. M.; Verzijl, G. K. M.; de
Vries, J. G. Angew. Chem., Int. Ed. 1998, 37, 662. (b) Stephan, M. S.; de
Vries, J. G. In Chemical Industries 82, Catalysis of Organic Reactions;
Ford, M. E., Ed.; Marcel Dekker Inc.: New York, 2000, p 379. For further
work on this concept see: Goossen, L. J.; Paetzold, J.; Ghosh, K. Synlett
2002, 1721 and references therein.
(6) Jeffery, T. In AdVances in Metal-Organic Chemistry; Liebeskind, L.
S., Ed.; JAI Press: Greenwich, CT, 1996; Vol 5, p 153. See also: Ziegler,
C. B., Jr.; Heck, R. F. J. Org. Chem. 1978, 43, 2941.
(7) The phrase “ligand-free” is used when no additional ligand, e.g., a
phosphine, is added to the reaction mixture to keep the palladium in solution.
(8) (a) Bumagin, N. A.; More, P. G.; Beletskaya, I. P. J. Organomet.
Chem. 1989, 371, 397. (b) Beletskaya, I. P. Pure Appl. Chem. 1997, 69,
471.
(2) de Vries, J. G. Can. J. Chem. 2001, 79, 1086.
(3) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 4176.
(4) van Strijdonck, G. P. F.; Boele, M. D. K.; Kamer, P. C. J.; de Vries,
J. G.; van Leeuwen, P. W. N. M. Eur. J. Inorg. Chem. 1999, 1073.
10.1021/ol035184b CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/15/2003

carboxylic acid derivatives.1f Recently, we developed a
practical ligand-free palladium recycle for Heck reactions
on aryl iodides, which can be used on-site in the same vessel
where the Heck reaction has been performed.9
lates.15 Optimized Jeffery conditions with use of MeNCy2
or HNCy2 as base in conjunction with Et4NCl also work in
the ligand-free palladium Heck reaction on aryl bromides,
but these reactions are rather slow.16
However, ligand-free approaches usually do not work for
the preferred aryl bromides.10 In most cases a precipitate of
palladium black forms in the early stages of the reaction and
the reaction usually stops well before full conversion. The
reason for this probably lies in the kinetics of these reactions.
Whereas with the aryl iodides olefin insertion is presumably
the rate-determining step, with aryl bromides it is most likely
the oxidative addition of this substrate to palladium that is
rate determining. This has great implications for the stability
of the resting state. Whereas with the aryl iodides most of
the palladium is in the form of palladium(II) complexes,9
with aryl bromides the resting state is most likely a less stable
palladium(0) complex.11 The Pd(0) species is subject to two
competing processes (Scheme 1). The catalyst can either
Whereas these ligand-free solutions function well they
defeat the purpose of our attempts to simplify the reaction
mixture. Based on the picture in Scheme 1 a more simple
solution presents itself. Whereas the Heck reaction presum-
ably is first order in palladium concentration, the aggregation
process must be higher order. Thus it would seem that by
lowering the catalyst-to-substrate ratio the point can be
reached where the oxidative addition outruns the aggregation
process. In practice, this concept works beautifully: use of
ligand-free Pd(OAc)2 in the range between 0.01 and 0.10
mol % leads to very acceptable reaction rates in the Heck
reaction of bromobenzene with butyl acrylate (NMP, NaOAc,
135 °C) (Figure 1).17,18 As shown in Figure 1 the reaction
Scheme 1. Fate of Palladium in Ligand-Free Heck Reactions
on Aryl Bromides
enter the catalytic cycle or aggregate to form initially soluble
palladium clusters, which at some point will turn into
insoluble palladium black. The latter process is self-catalyzed
and rapidly leads to the withdrawal of all palladium from
the catalytic cycle.
It has been shown that palladium clusters stabilized by
quaternary ammonium salts are active catalysts in the Heck
reaction on aryl bromides.12 Recently, several systems have
been described based on palladium particles stabilized by
coordinating polymers,13 dendrimers,14 or polyoxo-meta-
Figure 1. Effect of palladium/substrate ratio on the yield of the
Heck reaction between PhBr and n-butyl acrylate at 135 °C.
stops at an early stage at a palladium-to-substrate ratio of
1.28 mol %. In this reaction the presence of palladium black
is clearly visible. No palladium black was observed at the
lower concentrations, nor did the reaction stop halfway.
When the catalyst-to-substrate ratio is further lowered to
0.00125 mol % the catalysis still functions very well, but
the conversion is too slow to be practical.19 It is also clear
from Figure 1 that the turnover frequency increases going
(9) de Vries, A. H. M.; Parlevliet, F. J.; Schmieder-van de Vondervoort,
L.; Mommers, J. H. M.; Henderickx, H. J. W.; Walet, M. A. N.; de Vries,
J. G. AdV. Synth. Catal. 2002, 344, 996.
(10) Although aryl chlorides are cheaper than aryl bromides, the TON
numbers attained for aryl chlorides are much lower than those for aryl
bromides, so overall, a process based on an aryl bromide with low-loading
palladium is cheaper. See also the concluding paragraph in ref 3.
(11) As evidenced by the absence of Ar-Pd species in the MS (vide
infra). Also see: Jutand, A.; Negri, S.; de Vries, J. G. Eur. J. Inorg. Chem.
2002, 1711.
(15) Kogan, V.; Aizenshtat, Z.; Popovitz-Biro, R.; Neumann, R. Org.
Lett. 2002, 4, 3529.
(12) (a) Reetz, M. T.; Breinbauer, R.; Wanninger, K. Tetrahedron Lett.
1996, 37, 4499. (b) Beller, M.; Fischer, H.; Ku¨hlein, K.; Reisinger, C.-P.;
Herrmann, W. A. J. Organomet. Chem. 1996, 520, 257. (c) Reetz, M. T.;
Westermann, E. Angew. Chem., Int. Ed. 2000, 39, 165.
(13) (a) Klingelho¨fer, S.; Heitz, W.; Greiner, A.; Oestreich, S.; Fo¨rster,
S.; Antonietti, M. J. Am. Chem. Soc. 1997, 119, 10116. (b) Ley, S. V.;
Ramarao, C.; Gordon, R. S.; Holmes, A. B.; Morrison, A. J.; McConvey,
I. F.; Shirley, I. M.; Smith, S. C.; Smith, M. D. Chem. Commun. 2002,
1134. (c) Bergbreiter, D. E.; Osburn, P. L.; Li, C. Org. Lett. 2002, 4, 737.
(14) Rahim, E. H.; Kamounah, F. S.; Frederiksen, J.; Christense, J. B.
Nano Lett. 2001, 1, 499-501.
(16) Gu¨rtler, C.; Buchwald, S. L. Chem. Eur. J. 1999, 11, 3107.
(17) de Vries, A. H. M.; de Vries, J. G. WO 02/057199 to DSM nv,
2002.
(18) The low loading of ligand-free palladium (in our examples 0.01-
0.1 mol %) has been dubbed as homeopathic doses, a phrase also used by
Beletskaya and Cheprakov, see ref 1f, p 3009.
(19) Reetz has noted in his work on the use of dimethylglycine as
promotor for the Heck reaction that the promoting effect disappears at very
low palladium-to-substrate ratios, ligand-free Pd(OAc)2 at 0.009 mol %
sufficing for 77% conversion after 24 h. Reetz, M. T.; Westermann, E.;
Lohmer, R.; Lohmer, G. Tetrahedron Lett. 1998, 39, 8449.
3286
Org. Lett., Vol. 5, No. 18, 2003

from 0.08 to 0.00125 mol %.20 This can be explained by the
equilibrium between the palladium present in the clusters
and the palladium that is taking part in the catalytic cycle.
At lower concentrations this equilibrium is shifted away from
the clusters leading to a higher percentage of active catalyst.
To gauge the generality of this phenomenon and also to
establish the scope of the method we have screened a large
number of substrates in parallel, using the automated
synthesizer ASW 2000 of Chemspeed. Conversions and
selectivities were determined by GC-MS of samples of the
reaction mixtures taken after 1, 2, 5, and 15 h. (See Table 1
Figure 2. Observed selectivity in the ligand-free Heck reaction
for electron-poor olefins.
reactivity of the aryl bromide. Styrene also undergoes the
ligand-free Heck reaction smoothly (entries 21-24). The
selectivity of these reactions follows the general rules given
in Figure 2 and is similar to those found for ligand-containing
catalysts.1 Electron-rich olefins convert nicely too but show
the usual poor selectivity (entries 25-30),21 except for
3-buten-2-ol, which forms the ketone adduct in more than
90% (entries 31-32).
To benchmark the method, we compared our ligand-free
palladium system with the palladacycle developed by Herr-
mann, Beller, and co-workers22 (both at 0.02 mol % of
palladium). As shown in Figure 3, the two catalysts have a
Table 1. Heck Reactions with 0.05 mol % Pd(OAc)2 as a
Catalysta
time conv
entry
aryl bromide
olefin
(h)
(%)
1
2
3
4
5
6
7
8
9
PhBr
4-CHO-PhBr
4-CN-PhBr
4-CH3CO-PhBr
4-O2N-PhBr
2-CHO-PhBr
2-CN-PhBr
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
5
1
1
1
1
5
1
2
2
96
100
100
100
100
95
100
100
100
100
95
1-Br-naphthalene
9-Br-phenanthrene
10
11
12
13
2-Br,6-MeO-naphthalene n-Bu-acrylate
1
2
15
15
15
1
1
1
2
5
4-Br-biphenyl
2-Br-biphenyl
3-Br-pyridine
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
n-Bu-acrylate
t-Bu-acrylate
t-Bu-acrylate
t-Bu-acrylate
t-Bu-acrylate
t-Bu-acrylate
t-Bu-acrylate
styrene
94
100
100
100
100
100
99
91
92
99
99
14b 4-HO2C-PhBr
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
2-F-PhBr
3-F-PhBr
4-F-PhBr
4-Cl-PhBr
4-MeO-PhBr
4-Me2N-PhBr
PhBr
15
15
15
4-CH3CO-PhBr
styrene
2-Br,6-MeO-naphthalene styrene
15 >95
4-Cl-PhBr
PhBr
4-CH3CO-PhBr
2-Br,6-MeO-naphthalene CH2dCHOCyc
4-Cl-PhBr
styrene
CH2dCHOCyc
CH2dCHOCyc
15
15
15
15 >95
15 >95
99
85
99
CH2dCHOCyc
PhBr
N-vinyl-acetamide 15 >90
30d 4-CH3CO-PhBr
31
32
N-vinyl-acetamide
2
15
5
100
80e
99e
Figure 3. Rate comparison of catalysts in the reaction of PhBr
with n-butyl acrylate in NMP and NaOAc as base (140 °C).
PhBr
3-buten-2-ol
4-CH3CO-PhBr
3-buten-2-ol
a General conditions: 2 mmol of ArBr, 2.3 mmol of olefin, 0.05 mol %
of Pd(OAc)2, 2.4 mmol of NaOAc, NMP, 135 °C (total volume: 5 mL).
Conversion and selectivity (see Figure 2 and text) were determined by GC-
MS, using an internal standard. b 4.8 mmol of NaOAc. c Cy ) cyclohexyl.
d Octyl3N as base. e Selectivity of more than 90% to 1-aryl-3-butanone.
very similar kinetic profile in the Heck reaction between
bromobenzene and n-butyl acrylate.23
This intriguing result made us hypothesize that the active
catalyst in the palladacycle-catalyzed reaction also is a ligand-
(21) A mixture of a â-trans- and â-cis-arylated product was found, see
also: Andersson, C.-M.; Hallberg, A. J. Org. Chem. 1987, 52, 3529.
(22) (a) Herrmann, W. A.; Brossmer, C.; O¨ fele, K.; Reisinger, C.-P.;
Priermeier, T.; Beller, M.; Fischer, H. Angew. Chem., Int. Ed. Engl. 1995,
34, 1844. (b) Herrmann, W. A.; Brossmer, C.; Reisinger, C.-P.; Riermeier,
T. H.; O¨ fele, K.; Beller, M. Chem. Eur. J. 1997, 3, 1357. (c) Bo¨hm, V. P.
W.; Herrmann, W. A. Chem. Eur. J. 2001, 7, 4191.
and Figure 2.) A broad range of aryl bromides with electron-
withdrawing as well as electron-donating substituents showed
very high conversions with butyl acrylate (entries 1-20).
Reaction times range from 1 to 15 h depending on the
(20) TOF values at 2 h 40 min are the following: 309 (0.08 mol %),
(23) The observed TOF values for the palladacycle in NMP correspond
nicely with the ones reported in DMF, see ref 22.
787 (0.02mol %), 900 (0.00125 mol %) h-1
.
Org. Lett., Vol. 5, No. 18, 2003
3287

free palladium species. Recently, Nowotny et al. showed that
when using an immobilized palladacycle in a Heck reaction
the catalytic activity remained in the filtered solution,
whereas no activity remained in the solid.24 Similar findings
were reported by Rocaboy and Gladysz25 and by Beletskaya
et al.26 These authors refer to the intervention of nanoparticles
or clusters as catalysts in these Heck reactions. However,
the phenomenon of increasing TOF with decreasing catalyst
concentration, which seems to be generally associated with
the use of palladacycles as catalysts,25-27 can only be
explained by an equilibrium between a higher order pal-
ladium species and a monomeric species.28
activity at lower palladium concentrations is explained by a
preequilibrium between palladium clusters and monomeric
palladium that is contained in the catalytic cycle (Scheme
1), which shifts to the highly active monomeric palladium
at lower palladium concentrations.
Several ligand-free heterogeneous catalysts have been
reported for the Heck reaction on aryl bromides. These
include heterogeneous palladium on grafted molecular
sieves,32 zeolites,33 layered double hydroxides,34 hydroxy-
apatites,35 or just activated carbon or Al2O3.36 We assume
that these systems also work via a monomeric Pd(0) species
that may be in solution,36 where it becomes equivalent to
our method, or it may be contained in the framework of the
inorganic material.
In conclusion, Heck reactions on aryl bromides can
conveniently be performed with ligand-free palladium, added
as Pd(OAc)2 as long as the amount is kept as low as 0.01-
0.1 mol %. This procedure is more reliable than previous
methods using higher concentrations of palladium catalysts.
There is no need for extra added ligands or preformed
palladacycles; in fact most ligands retard the reaction. This
method is extremely attractive for large-scale production in
view of the low catalysts costs and the easy workup
procedure. The method has been scaled up by us to Kg size.
We next tried to confirm our suspicion through an ES-
MS study of the catalyst during the reaction. Indeed in both
-
reactions the only palladium species found was PdBr3 .
Neither in negative mode nor in positive mode did we detect
any phosphine-ligated palladium species during the reaction.
Mechanisms based on anionic species have been proposed
by Amatore and Jutand for phosphine-ligated species29 and
by Evans et al.30 and us9 for ligand-free palladium in Heck
reactions on aryl iodides. The MS experiment seems to
confirm the fact that in Heck reactions of aryl bromides
catalyzed by palladacycles the actual catalyst is the same
monomeric ligand-free palladium as in the ligand-free
system. The absence of arylated palladium species in the
MS, which is in contrast with the findings for aryl iodides,
seems to confirm that the oxidative addition is the rate-
Acknowledgment. We thank the Dutch Ministry of
Economic affairs for a subsidy under the EET Scheme (grant
nos. EETK97107 and EETK99104).
-
determining step. We presume that the PdBr3 is merely a
reservoir from which the Pd(0) is generated.31 The higher
Supporting Information Available: Experimental details
and ES-MS data on catalytic intermediates. This material is
available free of charge via the Internet at http://pubs.acs.org.
(24) Nowotny, M.; Hanefeld, U.; van Koningsfeld, H.; Maschmeyer, T.
Chem. Commun. 2000, 1877.
(25) (a) Rocaboy, C.; Gladysz, J. A. Org. Lett. 2002, 4, 1993. (b)
Rocaboy, C.; Gladysz, J. A. New J. Chem. 2003, 27, 39.
OL035184B
(26) Beletskaya, I. P.; Kashin, A. N.; Karlstedt, N. B.; Mitin, A. V.;
Cheprakov, A. V.; Kazankov, G. M. J. Organomet. Chem. 2001, 622, 89.
(27) (a) Albisson, D. A.; Bedford, R. B.; Scully, P. N. Tetrahedron Lett.
1998, 39, 9793. (b) Alonso, D. A.; Na´jera, C.; Pacheco, M. C. AdV. Synth.
Catal. 2002, 344, 172. (c) Gruber, A. S.; Zim, D.; Ebeling, G.; Monteiro,
A. L.; Dupont, J. Org. Lett. 2000, 2, 1287. (d) Iyer, S.; Ramesh, C.
Tetrahedron Lett. 2000, 41, 8981. (e) Gibson, S.; Foster, D. F.; Eastham,
G. R.; Tooze, R. P.; Cole-Hamilton, D. J. Chem. Commun. 2001, 779.
(28) Pfaltz and Blackmond, in their study on the kinetics of a Heck
reaction catalyzed by a palladacycle, assumed that this phenomenon is due
to a preequilibrium between a dimeric and a monomeric aryl-Pd(II) species.
They found half-order in [Pd]. See: Rosner, T.; Le Bars, J.; Pfaltz, A.;
Blackmond, D. G. J. Am. Chem. Soc. 2001, 123, 1848. For the cluster case
an order between 0 and 1 is expected.
(31) The actual mechanism is currently a subject of study. Very recently
a palladacycle was proposed as a reservoir of an active Pd(0) species, see:
Consorti, C. S.; Zanini, M. L.; Leal, S.; Ebeling, G.; Dupont, J. Org. Lett.
2003, 5, 983.
(32) Mehnert, C. P.; Weaver, D. W.; Ying, J. Y. J. Am. Chem. Soc. 1998,
120, 12289.
(33) Djakovitch, L.; Koehler, K. J. Am. Chem. Soc. 2001, 123, 5990
and references therein.
(34) Choudary, B.; Madhi, S.; Chowdari, N. S.; Kantam, M. L.; Sreedhar,
B. J. Am. Chem. Soc. 2002, 124, 14127 and references therein.
(35) Mori, K.; Yamaguchi, K.; Hara, T.; Mizugaki, T.; Ebitani, K.;
Kaneda, K. J. Am. Chem. Soc. 2002, 124, 11572.
(29) Amatore, C.; Jutand, A. Acc. Chem. Res. 2000, 33, 314.
(30) Evans, J.; O’Neill, L.; Kambhampati, V. L.; Rayner, G.; Turin, S.;
Genge, A.; Dent, A. J.; Neisius, T. J. Chem. Soc., Dalton Trans. 2002,
2207.
(36) (a) Biffis, A.; Zecca, M.; Basato, M. Eur. J. Inorg. Chem. 2001,
1131. (b) Davies, I. W.; Matty, L.; Hughes, D. L.; Reider, P. J. J. Am.
Chem. Soc. 2001, 123, 10139. (c) Ko¨hler, K.; Heidenreich, R. G.; Krauter,
J. G. E.; Pietsch, J. Chem. Eur. J. 2002, 8, 622.
3288
Org. Lett., Vol. 5, No. 18, 2003