Examination of ligand effects in the heck arylation reaction

Article abstract of DOI:10.1016/S0040-4020(00)00703-1

The Heck arylation reaction between aryl bromide and iodide with methyl acrylate was evaluated. A new class of ligands, R-N(CH₂CH₂PPh₂)₂, is compared with conventional monodentate and bidentate ligands, and was

Full text of DOI:10.1016/S0040-4020(00)00703-1

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 7975±7979
Examination of Ligand Effects in the Heck Arylation Reaction
*
È
Maryiam Qadir, Tobias Mochel and King Kuok (Mimi) Hii
Department of Chemistry, King's College London, Strand, London WC2R 2LS, United Kingdom
Received 22 May 2000; revised 19 July 2000; accepted 3 August 2000
AbstractÐThe Heck arylation reaction between aryl bromide and iodide with methyl acrylate was evaluated. A new class of ligands,
R±N(CH2CH2PPh2)2, is compared with conventional monodentate and bidentate ligands, and was found to confer greater catalyst stability.
The nitrogen substituent (R) has a noticeable effect on the rate of the turnover. We believe this is the ®rst report of the application of such
hemilabile ligands in Heck arylation chemistry. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
Results and Discussion
To appreciate the performance of these ligands in the Heck
arylation reaction, studies were carried out in parallel with
conventional monodentate and bidentate phosphine ligands,
such that direct comparison might be achieved.
The development of catalysts for the Heck arylation reac-
tion has been the subject of several recent high pro®le
reports, which describe the discovery of a number of palla-
dium metal complexes that are capable of achieving ex-
tremely high catalytic turnovers, even with fairly
unreactive aryl bromides and chlorides.1 Almost all of the
catalysts reported to date contain cyclometalated pallada-
cycles (Fig. 1).2±4 Although impressive turnovers are
attained, it can be envisaged that tuning the ligands
sterically and/or electronically will not be a trivial exercise.
Coupling between iodobenzene and methyl acrylate
Our ®rst chosen system was the catalysed reaction between
iodobenzene and methyl acrylate (Table 1). To our great
surprise, turnovers of about 1 million could be easily
achieved with `ligandless' catalysts such as palladium(0)
and palladium(II) complexes (entries 2 and 3). Furthermore,
As part of our recent studies on a new class of `tuneable'
phosphorus±nitrogen±phosphorus containing ligands, 4
(Fig. 2), we postulated that these ligands will be able to
stabilise fragile reactive intermediates via the hemilabile
nitrogen site, which allows the ligand to switch between
cis (PP) and trans (PNP) coordination modes easily without
necessitating the cleavage of a Pd±P bond (Fig. 3).5 If this is
true we expect the palladium metal complex to be more
stable in catalytic reactions, offering higher turnover
numbers without compromising reactivity. In this paper
we wish to compare the performance of these ligands with
conventional mono- and bi-dentate ligands towards Heck
arylation chemistry.
Figure 2. Different PNP ligands.
Figure 1. Recently reported catalysts for highly ef®cient Heck arylation reactions.
Keywords: C±C coupling; Heck arylation; homogeneous catalysis; palladium; P ligands.
* Corresponding author. Tel./fax: 144-020-78482810; e-mail: mimi.hii@kcl.ac.uk
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00703-1

7976
M. Qadir et al. / Tetrahedron 56 (2000) 7975±7979
dium precursors have limited turnovers of less than 250
(entries 1 and 2), whereas the addition of monodentate
phosphines seems to improve the catalyst's performance
somewhat (entries 3 and 4). The addition of bidentate
phosphines such as 1,2-bis(diphenylphosphino)ethane
(dppe), 1,3-bis(diphenylphosphino)propane (dppp) and
1,10-bis(diphenylphosphino)ferrocene (dppf) induce lower
turnovers compared to the monodentate PPh3 (entries 5±7).9
Figure 3. Different coordination modes adopted by PNP ligands.
Table 1. Effect of catalyst with/without phosphine ligands (reactions were
carried out in pressure tubes, with 1 mg catalyst, 3 equiv. of methylacrylate
with respect to iodobenzene, and 3 equiv. of NEt3)
para-Substituents on the rate of the Heck arylation
reaction
The relative rates offered by different para-substituted
monodentate phosphines are summarised in Table 3.
Percentage conversions were recorded after 7 h, from
which the comparative rates were calculated. The reactions
were then allowed to proceed for 31 h to determine the total
number of turnovers achievable by the individual systems.
From the results it is evident that the more electron-rich
triarylphosphines induce faster reactions (entries 3 and 4).
The ortho-substituted tris(methoxyphenyl)phosphine was
much slower than the para-analogue (entries 3 and 5),
implying an important steric effect on the rate of reaction.
When the relative rates were plotted against the Hammett
(spara) parameters, we obtained excellent linearity (Fig. 4).
This forms an interesting comparison to the results of Jutand
and Amatore,10 in which the rates of oxidative addition (one
of the possible rate-determining steps of the Heck catalytic
cycle) of aryl halides to palladium complexes coordinated
by triarylphosphines are reported to follow a bell-shaped
correlation.
Entry
Catalyst
Reaction time/h
% Conversion
TON
1
2
Pd(PPh3)45100
Pd2(dba)3.CHCl3
Pd(OAc)2
Pd(OAc)2/4a
Pd(OAc)2/PPh3
[(dppp)Pd(dba)]
109,106
4
4.5
19.5
19.5
6
100
95
97
100
100
100,000
95,000
97,000
100,000
100,000
3
4a
5
6
a 4b±4e showed the same reactivity.
the addition of certain ligands such as 1,3-bis(diphenylpho-
sphino)propane (dppp), PPh3 or 4a even appeared to slow
down the rates of these reactions (entries 4±6). Although
ligandless Heck arylation is nothing new,6,7 we believe this
provides direct evidence that the addition of phosphine
ligands is unnecessary, and may indeed have a decelerating
effect, on the Heck reactions involving aryl iodides.
Effect of using phosphorus±nitrogen±phosphorus
ligands, 4
Coupling reaction between 4-bromoacetophenone and
methyl acrylate8
In an earlier paper, we postulated that the introduction of a
hemilabile nitrogen donor between two phosphorus donor
atoms could improve the stability of palladium complexes
of diphosphine ligands.5 When ligands 4 were utilised in the
coupling reaction of aryl bromides, we ®nd that turnovers of
10,000 were achievable in all cases (Table 4).
In dramatic contrast, the reaction between methyl acrylate
and 4-bromoacetophenone is accelerated by the addition of
phosphine ligands (Table 2). Compared to the results
obtained with iodobenzene, this is quite a signi®cant obser-
vation as it implies that different rate-limiting steps operate
for Heck arylation reactions involving aryl iodides and
bromides.
The results show that the catalytic activity induced by the
PNP ligands is indeed better than monodentate and bi-
dentate diphosphine ligands. More importantly, the rate of
As shown in Table 2, the reactions with `ligandless' palla-
Table 2. Heck arylation of methyl acrylate with 4-bromoacetophenone using `common' phosphine ligands (reactions were carried out in pressure tubes.
Reagents: 3 equiv. of Et3N with respect to the aryl bromide and 3 equiv. of methylacrylate in anhydrous DMF)
Entry
Catalyst
Reaction time/h
Catalyst loading/mol%
% Conversion
TON
1
2
3
4
Pd(OAc)2
5
5
2
19
24
24
24
0.4
0.4
0.4
0.01
0.01
0.01
0.01
85
75
100
97
85
86
96
212.5
187.5
250
9,700
8,540
8,550
9,580
Pd2(dba)3.CHCl3
Pd(OAc)2/2 PPh3
Pd(OAc)2/2 PPh3
Pd(OAc)2/dppe
Pd(OAc)2/dppp
Pd(OAc)2/dppf
5a
6a
7a
a 3 equiv. of Na2CO3 as base in DMA.

M. Qadir et al. / Tetrahedron 56 (2000) 7975±7979
7977
Table 3. Effect of substituents on triarylphosphines (reactions were carried out in parallel in a reaction carousel. Reactions were left to run for 31 h to obtain
optimum turnover. Reagents: 0.01 mol% catalyst, 3 equiv. of Na2CO3 with respect to the aryl bromide and 3 equiv. of methylacrylate)
Entry
1
Ligand
% Conversion (reaction time/h)
Relative rates (krelative
)
45(7)
96(31)
1
2
3
4
40(7)
98(31)
0.878
1.251
1.798
56(7)
98(31)
81(7)
100(31)
5
25(7)
100(31)
0.545
Figure 4. Hammett correlation between reaction rates and para-substituent of the triarylphosphine.
Table 4. Reaction using PNP phosphine ligands (reactions were carried out in pressure tubes. Reagents: 0.01 mol% catalyst, 3 equiv. of Et3N with respect to the
aryl bromide and 3 equiv. of methylacrylate)
Entry
Ligand
Reaction time/h
Catalyst loading /mol%
% Conversion
TON
TOF/h21
1
2
3
4
5
4a
4b
4c
4d
4e
15
11
8
9
10
0.01
0.01
0.01
0.01
0.01
100
100
100
100
100
10,000
10,000
10,000
10,000
10,000
667
909
1,250
1,111
1,000

7978
M. Qadir et al. / Tetrahedron 56 (2000) 7975±7979
reaction is dependent on the nature of the nitrogen sub-
stituent, increasing in the order: tert-butyl (4c).n-butyl
(4d).benzyl (4e).p-anisole (4b).phenyl (4a). The trend
seems to ®t the electronic nature of the substituents, with the
most electron-donating group imparting twice the TOF
compared to the most electron-withdrawing group (entries
1 and 3). This is a very satisfying result as it con®rms our
theory that the presence of a hemilabile site could promote
both the stability and reactivity of the metal complex.
were mixed. Additional solvent was added to give a total
solvent volume of 2.5 mL, so as to achieve homogeneity
throughout the reaction. A small magnetic stirrer was
added, and the reaction vessel was sealed via the Te¯on
tap, before heating in an oil bath, with the reaction tempera-
ture carefully controlled by a digital thermostat (^18C). For
these reactions, sampling of reaction mixtures were
performed at regular intervals by cooling the pressure tube
down and extracting a small aliquot for GC/1H NMR
analyses.
Conclusion
Typical catalytic reaction with 4-bromoacetophenone
(10,000 turnovers): Under a dry nitrogen atmosphere,
Pd(OAc)2 (2.8 mg, 1.25£1022 mmol) and the ligand 4c
(6.2 mg, 1.25£1022 mmol) were dissolved in 10 mL of
DMA in a volumetric ¯ask, and the solution was gently
warmed to give a pale yellow solution. 4-Bromoaceto-
phenone (2.45 g, 12 mmol), methyl acrylate (3.3 mL,
36 mmol) and the appropriate base (36 mmol) were placed
in a Young's tube, to which 1 mL of the catalyst solution
was added, together with additional 1.5 mL of DMA (to
achieve homogeneity throughout the reaction). A small
magnetic stirrer bar was added, and the tube was then
sealed. The Young's tube was then heated in an oil bath,
its temperature carefully controlled by a digital contact
thermometer at 1408C (^18C). Analytical samples were
extracted as before.
The effect of different phosphine ligands has been studied
for the Heck arylation reaction. The reaction of methyl
acrylate with iodobenzene is not ligand accelerated,
whereas the reaction with bromoacetophenone is. The palla-
dium complexes of PNP ligands have also been examined,
and the nature of the hemilabile nitrogen donor is observed
to have a big effect on the turnover of the catalytic reaction.
All of these ligands are better catalysts than monodentate
and bidentate ligands in terms of greater catalyst stability
(all achieving 10,000 turnovers without losing activity).
Ligands 4c afforded the fastest turnover than any of the
ligands tested in this study.
Work is currently underway to prepare a second generation
of PNP ligands with further modi®cations, in the hope of
activating the metal complex for more dif®cult coupling
reactions with other aryl bromides and chlorides. Further
applications of these ligands in other transition-metal
catalysed reactions are also being sought.
Reaction carousel experiments: For comparison, the reac-
tions were also carried out in a Radley's 12-placed reaction
carousel on the same scale. Reaction mixtures were stirred
and re¯uxed under N2 atmosphere with a thermostated
(140^18C) heating block. Analytical samples of reaction
mixtures were extracted at regular intervals via a syringe
through a ®tted septum.
Experimental
The phosphines 4a±c were prepared according to
previously reported procedure.5 Ligands 4d±e were simi-
larly prepared in our laboratory recently.11 Iodobenzene
(Avocado) was passed through a column of neutral alumina
before use. All other phosphines and reagents were obtained
from commercial sources and used without any prior puri-
®cation. Solvents were degassed before use. Each catalytic
run was duplicated. Percentage conversion of the initial aryl
Acknowledgements
The authors thank ICI Strategic Research Funding for fund-
ing a summer research studentship (TM). K.K.H. especially
thanks Dr Bob Tooze (formerly of ICI Acrylics) for his
encouragement and support. Acknowledgements also to
Johnson Matthey for their generous loan of palladium salts.
1
halide was monitored by GLC and/or H NMR spectro-
scopy. When percentage conversion ,100, it signi®es that
the reaction did not progress beyond the stated turnover
number. TOF is only quoted for reactions that have achieved
100% conversion, and is thus an average value calculated by
dividing the time for complete conversion by the TON for
comparison of relative rates.
References
1. For reviews see: Crisp, G. T. Chem. Soc. Rev., 1998, 27, 427
and references therein.
È
2. (a) Herrmann, W. A.; Broûmer, C.; Ofele, K.; Reisinger, C. P.;
Pressure tube experiments
Riermeier, T.; Beller, M.; Fisher, H. Angew. Chem. Int. Ed. Engl.
1995, 34, 1844. (b) Herrmann, W. A.; Broûmer, C.; Reisinger,
È
C.-P.; Riermeier, T. H.; Ofele, K.; Beller, M. Chem. Eur. J.
1997, 3, 1357.
Typical catalytic reaction with iodobenzene: The appro-
priate palladium catalyst (1 mg) was carefully weighed out,
transferred into a 10 mL volumetric ¯ask and diluted to the
mark with the appropriate solvent. Dilutions were carried
out with the appropriate analytical pipettes and volumetric
¯asks as required. In a thick-walled tube ®tted with a Te¯on
Young's tap, iodobenzene (1.38 mL, 0.0123 mol), methyl
acrylate (3.3 mL, 0.036 mol), appropriate base and the
appropriate amount of diluted catalytic solution (1 mL)
3. Ohff, M.; Ohff, A.; Milstein, D. Chem. Commun. 1999, 357.
4. Albisson, D. A.; Bedford, R. B.; Scully, P. N. Tetrahedron Lett.
1998, 39, 9793.
5. Hii, K. K.; Thornton-Pett, M.; Jutand, A.; Tooze, R. P.
Organometallics 1999, 18, 1887.
6. Jeffrey, T.; David, M. Tetrahedron Lett., 1998, 39, 5751 and
references therein.

M. Qadir et al. / Tetrahedron 56 (2000) 7975±7979
7979
7. Reetz, M. T.; Westermann, E.; Lohmer, R.; Lohmer, G.
Tetrahedron. Lett. 1998, 39, 8449.
Heck arylation reactions see: Cabri, W.; Candiani, I.; Bedeschi,
A.; Santi, R. J. Org. Chem., 1992, 57 3558.
8. The reactions with aryl bromides are substantially slower.
To avoid excessive reaction times, the turnover number
was reduced (maximum of 10,000) by increasing catalyst
loading.
10. (a) Amatore, C.; Jutand, A.; Meyer, G. Inorg. Chim. Acta
1998, 273, 76. (b) Amatore, C.; Jutand, A. Coord. Chem. Rev.
1998, 180, 511.
11. Rahman, M. S.; Steed, J. W.; Hii, K. K. Synthesis 2000, 1320.
9. For the effect of bidentate ligands on the regioselectivity of