Use of a bulky phosphine of weak σ-donicity with palladium as a versatile and highly-active catalytic system: Allylation and arylation coupling reactions at 10-1-10-4 mol% catalyst loadings of ferrocenyl bis(difurylphosphine)/Pd

Article abstract of DOI:10.1016/j.tet.2005.06.065

Carbon-carbon(sp²-sp² and sp¹-sp ²) and carbon-nitrogen (nucleophilic allylation) coupling processes are promoted by a catalytic system containing [PdCl(η³-C ₃H₅)]₂ with the new ferrocenyl bis(difurylphosphine) 1,1′-bis[di(5-methyl-2-furyl)phosphino]ferrocene, Fc[P(Fu^Me)₂]₂. Starting from aryl bromides or allylic acetates this versatile catalyst system may be used at low palladium loadings (10^-1-10^-4 mol%) in some Heck, Suzuki, Sonogashira and allylic amination reactions to give cross-coupled products in excellent yield. Remarkably high activity is obtained in allylic substitution reactions, providing a significant impetus for the development of bulky phosphines possessing weak σ-donicity for this particular reaction.

Full text of DOI:10.1016/j.tet.2005.06.065

Tetrahedron 61 (2005) 9759–9766
Use of a bulky phosphine of weak s-donicity with palladium as a
versatile and highly-active catalytic system: allylation and
arylation coupling reactions at 10K1–10K4 mol% catalyst loadings
of ferrocenyl bis(difurylphosphine)/Pd
a
a
a
b,
Jean-Cyrille Hierso,a, Aziz Fihri, Regine Amardeil, Philippe Meunier, Henri Doucet
*
*
´
and Maurice Santellib
a
` ` ´ ´
Laboratoire de Synthese et Electrosynthese Organometalliques UMR-CNRS 5188, Universite de Bourgogne, 9 avenue Alain Savary,
21078 Dijon, France
` ´
Laboratoire de Synthese Organique UMR-CNRS 6180, Faculte des Sciences de Saint Jerome, Universite Aix-Marseille,
b
´ ˆ
´
Avenue Escadrille Normandie-Niemen, 13397 Marseille, France
Received 11 April 2005; revised 10 May 2005; accepted 17 June 2005
Available online 18 July 2005
Abstract—Carbon–carbon(sp2–sp2 and sp1–sp2) and carbon–nitrogen (nucleophilic allylation) coupling processes are promoted by a
catalytic system containing [PdCl(h3–C3H5)]2 with the new ferrocenyl bis(difurylphosphine) 1,10-bis[di(5-methyl-2-furyl)phosphino]
ferrocene, Fc[P(FuMe)2]2. Starting from aryl bromides or allylic acetates this versatile catalyst system may be used at low palladium loadings
(10K1–10K4 mol%) in some Heck, Suzuki, Sonogashira and allylic amination reactions to give cross-coupled products in excellent yield.
Remarkably high activity is obtained in allylic substitution reactions, providing a significant impetus for the development of bulky
phosphines possessing weak s-donicity for this particular reaction.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
catalyst loadings impose financial constraints on scaling up
reactions, or at least in problems associated with catalyst
removal. The necessity to provide catalytic systems
minimising the consumption of expensive transition-metals
can be fulfilled through the synthesis of accessible, efficient
and inexpensive ligands. The new bis(difurylphosphino)-
ferrocene ligand 1 presented herein was developed
following this objective, and pertains to a family of highly
efficient catalytic auxiliaries based on ferrocenyl poly-
phosphines, which we have synthesized (2a and 2b
Scheme 1).3–6
Palladium-catalysed carbon–carbon and carbon–heteroatom
bond formations represent important catalytic processes
with extensive scope in contemporary synthetic chemistry.
The progress made since the late 1960’s, following the
seminal studies by Heck, Sonogashira, Suzuki, Tsuji and
others,1 provides a variety of protocols for the production of
organic building blocks of prodigious interest in material
science, medicine or agronomy. Some practical problems
exist, which will require addressing for these reactions to
move from academic laboratories to industrial processes.2
Indeed, until very recently reported catalytic systems were
most often used at 1–10 mol% palladium loadings at best,
which is non-efficient in terms of turnover number and
sometimes of turnover frequency. Concomitantly, such high
Keywords: Catalysis; Palladium; Ferrocenylphosphine; Furyl; Allylic
substitution; Cross-coupling; Ligand effect.
*
Corresponding authors. Tel.: C33 380 39 6106, fax: C33 380 39 3682
(J.-C.H.); tel.: C33 491 28 8416; fax: C33 491 98 3865 (H.D.);
e-mail addresses: jean-cyrille.hierso@u-bourgogne.fr;
henri.doucet@univ.u-3mrs.fr
Scheme 1.
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.06.065

9760
J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
2. Results and discussion
reported for the parent compound: 1,10-bis[di(-2-furyl)-
phosphino]ferrocene, available by reaction of bis(dichloro-
phosphino)ferrocene with 2-lithiofuran.23
2.1. Preparation and characterization of the ligand 1,10-
bis[di(5-methyl-2-furyl)phosphino]ferrocene
Table 1 summarizes some of the 31P NMR chemical shifts
recorded for ferrocenyl heteroannular diphosphines, includ-
ing the data for 1, which displays a noticeable high-field
signal. The electronic difference expected between the
phosphorus atoms of 1 and the related ferrocenyl diphos-
phines incorporating aryl or alkyl substituents was
confirmed. Evidence concerning the intrinsic low Lewis-
basicity of phosphorus bearing 2-furyl substituents have
been collected and reported.27
A literature survey from the last 20 years concerning
palladium-catalysed cross-coupling reactions clearly shows
that much progress in this field was accomplished from the
synthesis/identification of efficient auxiliary ligands. Such
species not only allowed a wide range of synthetic advances
but moreover, provided interesting mechanistic investi-
gations. Amongst these auxiliaries, bulky phosphines with
high s-donicity emerged as powerful ligands.7–15 Their
efficiency was initially surprising since strong s-donor
ligands would not be expected to facilitate the crucial
transmetalation16 and reductive elimination steps,17 but
rather would be assumed to be powerful inhibitors. These
results indicated the importance associated with the often
rate-limiting oxidative-addition step, particularly with
substrates containing strong carbon-halide bonds. Addition-
ally, the success of bulky phosphines has been suggested to
be derived from the ability of these ligands to promote
geometrically driven coordinative unsaturation and
palladium reactivity (mainly due to steric hindrance2,18,19
and/or cone/bite angle20). These geometric features appear
to be important to transmetalation2 and reductive elimin-
ation21 steps, when utilising strongly electron-donating
ligands. Finally, investigations on ligands exhibiting both
pronounced electronic and steric properties, contribute to
synthetic advances, as well as to mechanistic understanding.
Table 1. 31P NMR chemical shifts for ferrocenyl heteroannular
diphosphines
Fe[h5-C5H4(PR2)]2
31P NMR (ppm)/solvent
RZt-Bu
RZi-Pr
RZCy
RZPh
RZMenthyl
27.114/C6D6
K0.225/C6D6
K9.726/CDCl3
K17.228/CDCl3
K24.125/CDCl3
K25.225/C6D6
K63.8/CDCl3
K64.923/CDCl3
RZ5-Me-2-furyl
RZ2-Furyl
The molecular structure of Fc[P(FuMe)2]2 shows that in the
solid state the cyclopentadienyl rings adopt an eclipsed
conformation (Fig. 1). In contrast to dppf (dppfZ1,10-
bis[diphenylphosphino]ferrocene), for which the phos-
phorus atoms point away from each other (Cp1–CNT1–
CNT2–Cp2 torsion angle 1808, antiperiplanar staggered
conformation), in 1 a 728 only torsion angle is observed for
Cp1–CNT1–CNT2–Cp2 (synclinal eclipsed).29 We did not
detect any noticeable ordered crystalline arrangement or
hydrogen bonding, which would explain such a geometry.
Bond angles and distances are within the expected values.
As a consequence of the complexity in the prediction of a
rate-limiting step,20 it did not appear unreasonable to us to
test, in more depth, a range of cross-coupling reactions
using a comparatively ‘forgotten’ strongly electron-with-
drawing bulky phosphine ligand.22 We choose to build a
weakly s-donor phosphine on the robustness of the
ferrocenic backbone.3 Ferrocenic phosphines with phos-
phorus bearing heterocyclic substituents23 are notably less
developed than the corresponding ferrocenyl diphosphines
bearing aryl or alkyl substituents.14,24–26 Nevertheless, the
new ligand 1,10-bis[di(5-methyl-2-furyl)phosphino]ferro-
cene, 1, is obtained pure in 80% isolated yield from reaction
of lithiated ferrocene with the bis(5-methyl-2-furyl)bromo-
phosphine BrP(FuMe
)
(Scheme 2). Compound 1 was
2
isolated in pure form on a 10 g scale, as a crystalline orange
powder, insensitive to air and moisture, easy to handle and
crystallise. This synthesis is different to that previously
˚
Figure 1. Molecular structure of 1. Selected bond lengths (A) and angles
(8): Fe(1)–CNT(1) 1.655(2), P(1)–C(5) 1.812(2), P(1)–C(6) 1.804(2), P(1)–
C(11) 1.803(2), O(1)–C(6) 1.385(3), O(2)–C(11) 1.385(3), CNT(1)–Fe(1)–
CNT(2) 178.18(14), C(11)–P(1)–C(5) 103.80(10), C(11)–P(1)–C(6)
102.77(10), O(1)–C(6)–P(1) 121.77(16), O(2)–C(11)–P(1) 103.80(10).
Scheme 2.

J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
9761
2.2. Catalytic results in ‘Suzuki’, ‘Heck’ and
‘Sonogashira’ cross-coupling reactions
reaction were conducted using 4-bromoanisole with n-butyl
acrylate without addition of tetraalkyl ammonium salts,
which are often used to delay palladium black formation.
Here, excellent conversions were obtained in the presence of
10K1 and 10K2 mol% catalyst (entries 5–6, TON 10,000),
but lowering the concentration inhibits completely the
reaction. At this stage, the results obtained without any
further optimization,30 that meet the definition of high-
turnover catalysts (HTC) proposed by Farina2 (a catalyst
that can lead to quantitative conversion of starting materials
at a load of 0.1 mol% or less), however did not completely
satisfy our activity criteria for the Suzuki–Miyaura cross-
coupling and Heck alkenylation reaction.
To evaluate the ligand properties in cross-coupling reactions
systematically, we first tested the performance of 1 in
Suzuki–Miyaura cross-coupling reactions using the
activated substrate 4-bromoacetophenone with phenyl-
boronic acid, following previously reported procedures.3
The reaction of aryl halide (10 mmol), aryl boronic acid
(20 mmol) at 130 8C during 20 h in dry xylene or DMF,
under argon, affords the corresponding products in presence
of palladium/phosphine catalysts and K2CO3. Surprisingly,
regarding the low basicity of phosphorus, a high turnover
number (TON) of 100,000 with a complete conversion was
obtained in presence of 10K3 mol% catalyst (Table 2, entry
1), a result equivalent to that obtained with the more
electron-rich tetraphosphine 2a under identical con-
ditions.3 A conversion of 18% is obtained in the presence
of 10K4 mol% catalyst (TON 180,000, entry 2) after 20 h.
Nevertheless, this facile coupling-reaction is to be used as a
guide and not as a stern benchmark to screen new Pd-
catalysts.9 We tested a more demanding substrate, the
deactivated and electron-rich organohalide, 4-bromo-
anisole. Complete conversion was obtained in the presence
of 10K2 mol% catalyst (TON 10,000, entry 3), but
disappointingly a lower concentration produces no cross-
coupled product (entry 4). Under identical conditions the
tetraphosphine 2a facilitated a TON of 77,000, in the
presence of 10K3 mol% catalyst.3
We next, compared the activity of 1 in the Sonogashira
arylation of phenylacetylene with the activated and
deactivated substrates: 4-bromoacetophenone and 4-bromo-
anisole, respectively. Using 4-bromoacetophenone in the
presence of 10K4 mol% catalyst, a TON of 920,000, among
the highest reported,4,31 was obtained (Table 3, entry 10).
4-Bromoanisole, as expected was quantitatively converted
(at 10K1 mol% catalyst concentration) at a lower TON of
1000. Note that under the same reactions conditions (i.e.,
using K2CO3 as base and 10K1 mol% catalyst) the absence
of ligand lead to no conversion.31 In the presence of a large
excess of PPh3, 5% conversion was observed, 3% with
P(o-tol)3 and 50% maximum using an excess of 1,4-
bis(diphenylphosphino)butane (dppb). With a more
demanding substrate such as an activated aryl chloride:
4-chloroacetophenone; no coupling product was obtained,
even at 0.4 mol% catalyst loading, whereas under the same
The preliminary experiments on the Heck alkenylation
Table 2. Suzuki and Heck reactions catalysed by the Fc[P(FuMe)2]2/palladium system
Entry
Aryl bromide
Aryl boronic acid or alkene
PhB(OH)2
Ratio substrate/catalyst
Yield%
1
2
3
4
5
6
7
4-MeCOC6H4Br
100,000
1,000,000
10,000
100,000
1000
10,000
100
18
100
0
100
99
0
4-MeOC6H4Br
4-MeOC6H4Br
PhB(OH)2
nBuOCOCH]CH2
100,000
Conditions: catalyst [PdCl(h3–C3H5)]2/1: 1⁄2. Suzuki: ArBr 1 equiv, ArB(OH)2 2 equiv, K2CO3 2 equiv, 20 h, 130 8C, xylene. Heck: ArBr 1 equiv, alkene
2 equiv, K2CO3 2 equiv, 20 h, 130 8C, dimethylformamide (DMF).
Table 3. Sonogashira reaction catalysed by the Fc[P(FuMe)2]2/palladium system
Entry
Aryl halide
Alkyne
Ratio substrate/catalyst
Yield%
8
9
10
11
12
13
4-MeCOC6H4Br
PhC^CH
10,000
100,000
1,000,000
250
1000
10,000
100
100
92
100
100
Traces
4-MeOC6H4Br
PhC^CH
Conditions: catalyst [PdCl(h3-C3H5)]2/1: 1⁄2. ArBr 1 equiv, alkyne 2 equiv, K2CO3 2 equiv, CuI 0.05 equiv, 20 h, 130 8C, DMF.

9762
J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
reactions conditions the triphosphine ligand 2b leaded to
82% conversion.4
sophisticated ligands and of polluting additives) is very
limited, the reported examples being rare.34–36
In the reactions conducted with 1, very high TONs were
obtained when activated substrates were used, and
comparatively disappointing results were observed from
more demanding aryl halides. From these results in C–C
cross-coupling reactions, several conclusive trends, which
are discussed in Section 2.4, led us to focus our further
catalytic investigations on the allylic amination reaction.
The results obtained in allylic amination with various allylic
acetate and amine substrates are summarised in Table 4. To
evaluate the scope and limitation of the system Pd/1, allyl
acetate was reacted with primary and secondary aliphatic
and cyclic amines of various steric and nucleophilic
features. We were delighted to see that without optimisation
procedures our system allowed to reach the best turnover
frequencies (TOFs) reported to date, and this at rt, in the
absence of base37 or any additive and in the presence of as
little as 0.01 mol% catalyst. A complete conversion was
obtained from the primary amine aniline with a satisfactory
selectivity of 96% in monoallylaniline (Table 4 entry 14,
TOF 10,000 hK1).38 An excellent 98% conversion was also
achieved from the cyclic secondary amine pyrrolidine (entry
15, TOF 4900); we verified the facile access to the product
(pure) after classical work-up procedures, with an isolated
yield of 88%. Interestingly, only a slightly decreased 85%
conversion was obtained with the less nucleophilic cyclic
morpholine (entry 16, TOF 4250 hK1) under the same
conditions; morpholine can be quantitatively converted in
3 h at rt. The coupling of primary amine benzylamine
appeared more difficult and less selective (entry 17, TOF
1825 hK1 for a 75% selectivity in monoallylation product).
In addition to primary amines and cyclic secondary amines,
also more challenging aliphatic secondary amines such as
n-dioctylamine, can be coupled efficiently. A complete
conversion of the sterically demanding diisopropylamine
was obtained at 80 8C in the presence of 0.1 mol% catalyst
2.3. Catalytic results in ‘Tsuji-Trost’ allylic amination
Allylamines are important building blocks in modern
chemistry and their preparation is a key synthetic and
industrial goal.32 The allylamine fragment is found in many
natural products, and can undergo a range of different
chemical processes on the C]C bond (functionalization,
oxidation, etc.). One of the most attractive and preferred
ways to form allylamines is by palladium-catalysed
nucleophilic allylic substitution, in the presence of a
stabilising ligand, mostly phosphines. In this field, a large
number of research groups have mainly focused on the
development of chiral catalysts, inducing enantioselectivity
in allylic alkylation reactions in general. Some of these
complexes have been applied to allylic amination.32,33
Consequently, even in the achiral form of the reaction, the
catalytic systems reported, to date, are generally used at
1–5 mol%. The knowledge related to more efficient
catalytic systems in terms of scope and catalyst economy
(reduction of the loading of the expensive catalytic metal, of
Table 4. Allylic amination reaction catalysed by the Fc[P(FuMe)2]21/palladium system
Entry
Allylic acetate
Amine
Ratio substrate/
catalyst
Reaction conditions
Yield%
Selectivity %
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Allyl acetate
Allyl acetate
Allyl acetate
Allyl acetate
Allyl acetate
H2NPh
10,000
10,000
10,000
10,000
1000
10,000
100,000
1000
1000
10,000
1000
10,000
1000
rt,a 1 h
rt,a 2 h
rt,a 2 h
rt,a 4 h
rt, 20 h
100
98
85
73
100
45
5
96
100
76
100
26
100
48
100
76
96/4 (mono/di)
100
100
75/25 (mono/di)
100
100
100
100
94/6 (lin/brch)
94/6
94/6 (lin/brch)
94/6
93/7 (lin/brch)
93/7
99/1 (lin/brch)
99/1
94/6 (lin/brch)
94/6 (lin/brch)
100
–(CH2)4HN–
–[O(CH2)4]HN–
H2NCH2Ph
HN(n-octyl)2
Allyl acetate
3-Phenylallyl acetate
HN(i-propyl)2
HNEt2
80 8C,a 2 h
rt, 20 h
3-Phenylallyl acetate
3-Phenylallyl acetate
E-Hex-2-en-1-yl acetate
–(CH2)4HN–
–[O(CH2)4]HN–
HNEt2
50 8C, 20 h
50 8C, 20 h
50 8C, 20 h
10,000
250
1000
1000
1000
E-Hex-2-en-1-yl acetate
E-Hex-2-en-1-yl acetate
E-Hex-2-en-1-yl acetate
–(CH2)4HN–
–[O(CH2)4]HN–
HN(n-octyl)2
50 8C, 20 h
50 8C, 20 h
50 8C, 20 h
100
98
59
11
250
1000
100
Conditions: catalyst [PdCl(h3-C3H5)]2/1: 1⁄2. No base. Allylic acetate 1 equiv, amine 2 equiv, tetrahydrofuran.
a Carried out in toluene. For 100% conversion in 20 h the reaction time was not minimised.

J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
9763
in 2 h (entry 21, TOF 480 hK1),36 and of the long alkyl
chain-bearing dioctylamine at rt after a longer period
(entries 18–20, best TOF 250 hK1).
multiply the rates of transmetalation and reductive
elimination, steps which would become determining.41
This hypothesis is consistent with the early results from
Farina and Krishnan,16 on the use of trifurylphosphine and
triphenylarsine (ligands of weaker basicity compared to
PPh3) in the Stille coupling reaction with substrates capable
of undergoing fast oxidative addition (aryl iodides for
instance); the Stille coupling of organostannanes being a
reaction, for which transmetalation is decisive and possibly
rate-determining.42–44
This new catalytic system was also found to be remarkably
active and selective for the more difficult amination of
substituted allylic acetates. Using the same pool of amines,
we focused our catalytic investigations on coupling reaction
of 3-phenylallyl acetate and E-hex-2-en-1-yl acetate at 0.1
and 0.01 mol% Pd/1, at moderate temperature on a short
period of time. Starting from diethyl amine, pyrrolidine and
morpholine a complete conversion was obtained from
0.1 mol% catalyst (entries 22, 24 and 26). Lowering the
catalyst loading to 0.01 mol% resulted in TONs of 7600,
2600 and 4800, respectively (entries 23, 25 and 27). In each
case, good regioselectivity was observed for the linear
product (93–94%), especially concerning the primary
amine: HNEt2; for which monoallylation occurs exclusively
(for primary amines, theoretically five different products
could be obtained: linear/branched mono or diallylic
product combinations). The selectivity is even higher in
linear monoallylamine for the coupling of HNEt2 to the
E-hex-2-en-1-yl acetate (99%, entries 28, 29). For the cyclic
secondary amines these coupling reactions proceeded in
O90% yield using as little as 0.1 mol% catalyst (94%
selectivity in linear product, entries 30, 31). Finally, only in
the course of the addition of dioctylamine, lower TONs and
TOFs were observed, since to reach 60% conversion in 20 h
at 50 8C, 0.4 mol% catalyst was required (entries 32, 33).
Nevertheless, the present tendency in cross-coupling
reactions to use the cheaper aryl chlorides (very reluctant
to oxidative addition)45,46with a view to industrial appli-
cations, leaves only a small place to the development of
weakly s-donating ligands,47 except if other strategies are
proposed like the systems developed by Bedford using
mixed ligand systems such as bulky s-acceptor/s-donor
ligands (Pd/{di-tert-butylphenyl}phosphite/tricyclohexyl-
phosphine).48,49
The tendencies, discussed above, are somewhat confirmed
by the results obtained in Heck alkenylation and
Sonogashira cross-coupling reactions. The system Pd/1
gave a TON of 104 for the Heck reaction of the deactivated
4-bromoanisole and n-butylacrylate, while the Pd/triaryl-
phosphite system reported by Albisson et al. gave complete
conversion50 at a TON of 440–500 (again for less
demanding substrates TONs above 106 are reported).
Finally, consistent with our hypothesis of the changes of
the rate determining-step, depending on the substrate, an
extremely high TON of 920,000 (with almost complete
conversion) was obtained for 4-bromoacetophenone
(facile oxidative addition, the rate-determining step being
reductive elimination or transmetalation), while when
4-bromoanisole was used a TON of 102 was reached (rate-
determining step is thus, oxidative addition). Clearly these
hypotheses require further comparative kinetics studies to
be confirmed.
2.4. Discussion
To the best of our knowledge, only a few reports exist on the
use of HTC incorporating very poor s-donating phosphine
ligands in Suzuki–Miyaura cross-coupling reactions.22,39,40
Our experiments from coupling aryl bromide with phenyl
boronic acid, confirmed the early investigations by
Albisson et al. using a system Pd/triaryl phosphite catalyst
system.39,40 Under the reactions conditions, which are rather
close to those used herein (toluene, 110 8C, K2CO3 as a
base), the authors obtained complete conversion of the
activated 4-bromoacetophenone with a turnover number of
106, while the complete conversion for the deactivated
4-bromoanisole was obtained at a 103 TON. Whereas
comparisons between different systems must be treated with
some caution, it is worth noting that for the same reactions,
we obtained complete conversion from 4-bromoaceto-
phenone with a TON of 105 and complete conversion
from bromoanisole at 104 TON. Under similar conditions,
the Pd/1 catalyst combination shows an order of magnitude
weaker activity for the electron-rich substrate and an order
of magnitude higher activity for the electron-poor substrate.
Our system, as in the Pd/triaryl phosphite combination,
exhibits no propensity to couple even electronically-
activated aryl chlorides such as 4-chloroacetophenone. A
rather simple explanation of this noticeable parallel
behaviour could be linked to the subtle changes depending
on the substrate of the rate determining-step: thus, for more
demanding substrates, the oxidative addition is rate-
determining (and severely disfavoured by poor s-donating
ligands). In contrast, for strongly activated substrates (aryl
iodides and activated bromides) where the oxidative
addition is comparatively facile, the same ligands would
To demonstrate the potential and catalytic utility of bulky
Scheme 3.

9764
J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
phosphines of weak s-donicity we choose to apply the Pd/1
catalytic system to a reaction where the possible
determining-step required an electrophilic metallic tran-
sition-state: for instance the palladium-catalysed allylic
amination with the nucleophilic attack on the allyl ligand
(Scheme 3).51
an inert atmosphere of argon using conventional vacuum-
line and glasswork techniques. Solvents were dried and
freshly distilled under argon. Reagents were purchased from
commercial suppliers. CDCl3 and CD2Cl2 degassed and
stored over molecular sieves under argon were used for
NMR studies. Xylene and DMF analytical grade (98%)
were not distilled before use. Potassium carbonate (99C%)
was used without drying. Except for the diphosphine ligand
and for some of the aryl halides and amines, which were
distilled before use, the organic and organometallic products
were received from commercial sources, and used without
further purification. Flash chromatography was performed
The recent mechanistic studies on allylic amination by
Kuhn and Mayr provided meaningful kinetic evidence.52
Indeed, in this relevant work, the use of various amines
(diethylamine, piperidine) and 3-phenylallyl acetate
(cinnamyl acetate) in presence of either PPh3 or the less
electron-donor P(OPh)3 conducted to global reaction rates
two orders of magnitude higher in favour of P(OPh)3. The
excellent results obtained with our system could be
rationalized through the catalytic cycle depicted in
Scheme 3. Relevant reports discuss the general rate-
determining step in the soft nucleophilic allylic substi-
tutions.51–57 We propose that, as anticipated, that the high
TOFs and good TONs reported herein are the result of the
ability of the ligand 1 to electronically increase the
electrophilicity of the Pd(II)/allyl transient species
(accelerating the nucleophilic attack, which seems to be
the rate-determining step) and to sterically stabilize the
Pd(0) complex resulting from product elimination. This is in
1
on silica gel (230–400 mesh). Elemental analyses, H, 31P
and 13C NMR were performed in our laboratories (on
Bruker 300). The evolution of catalyzed reactions was
followed by GC (or GC/MS) and NMR for high boiling
point substrates, and by GC for low boiling point substrates.
For the X-ray diffraction structure of 1, data were collected
on a Nonius Kappa CDD (Mo Ka) diffractometer at 110 K.
The structure was solved by a Patterson search program and
refined by full-matrix least-squares methods based on F2
using SHELX97 with the help of the WinGX program at the
´
‘Universite de Bourgogne’.
4.2. Synthesis of 1,10-bis[di(5-methyl-2-furyl)phosphino]
ferrocene, Fc[P(FuMe)2]2, 1
˚
agreement with the early work of Akermark and
co-workers.51
The starting products BrP(FuMe)2 and FeCp2Li2-TMEDA
were synthesised in high yield following procedures
described in the literature.58,59 To a stirred suspension of
6.78 g (21.61 mmol) of FeCp2Li2-TMEDA in 60 mL of
hexane was added dropwise a solution of 11.8 g
3. Conclusion
In summary, the synthesis, characterisation and catalytic
behaviour of the new bulky ferrocenic ligand of low
s-donicity, 1,10-bis[di(5-methyl-2-furyl)phosphino]ferro-
cene 1, has been reported. In cross-coupling reactions and
Heck alkenylation processes starting from aryl bromides,
the system using Pd/1 allows high turnover catalysis
(%0.1 mol% catalyst) since a total conversion is obtained
for the model substrates: 4-bromoacetophenone and
4-bromoanisole. However, more demanding substrates
(aryl chlorides) are not turned-over in the presence of low
concentrations of catalyst. In contrast, the ligand is revealed
to be among one of the more efficient auxiliaries for
palladium-catalysed allylic substitution processes. Facile
and convenient reaction conditions were employed in allylic
amination, since most of the screened substitution reactions
were conducted: (i) from acetates in the absence of base
(neutral conditions);37 (ii) at 20 or 50 8C temperatures; (iii)
in less than 24 h; (iv) in the presence of 0.1–0.01 mol%
catalyst. Thus, a subtle combination of electronic and steric
properties of the ligand produces a highly active catalyst for
allylic amination. In particular, the ligand is expected to
substantially enhance the rate of nucleophilic attack on the
palladium(II)allyl species. Further studies will address and
enlarge the scope of the present system in allylic
substitutions.
(43.22 mmol) BrP(FuMe
)
in 25 mL hexane. After 2 h
2
stirring at rt, 20 mL of degassed water was added. An
orange powder was obtained, which was washed three times
with hexane and then dried under vacuum. Purification was
carried out by dissolution in chloroform and filtration on
silica gel. 9.86 g of 1 were obtained (17.30 mmol, yield
80%). The product was recrystallised from a chloroform/
hexane mixture.
4.2.1. Spectroscopic and single X-ray characterization of
1
the ligand Fc[P(FuMe)2]2. H NMR (CDCl3, 300.13 MHz,
d in ppm): dZ2.33 (s,12H, MeFu); 4.15 and 4.30 (m, 8H, Cp)
5.96 and 6.52 (m, 4H, HFu); 31P {H} NMR (CDCl3,
121.49 MHz): dZK63.8 (s); 13C {1H} NMR (CD2Cl2,
2
75.47 MHz): dZ12.95 (s, 4C, MeFu), 71.02 (d, 4C, JCP
Z
5.4 Hz, CpCH), 73.18 (d, 4C, 3JCPZ17.2 Hz, CpCH), 73.02
(d, 2C,1JCPZ3.6 Hz, Cp), 105.80 (d, 4C, 2JCPZ5.7 Hz, C]
C(O)Me), 119.75 {d, 4C, JCPZ21.7 Hz, sp2 CH]C(O)},
3
1
149.45 (d, 4C, JCPZ4.1 Hz, P–C(O)]C), 155.47 (d, 4C,
4JCPZ2.3 Hz, Me–C(O)]C). Anal. Calcd. for
C30H28FeP2O4: C 63.17, H 5.20. Found C 63.05, H 4.95.
Crystal data for 1. C30H28FeO4P2, MZ570.32, ortho-
˚
rhombic, space group Pbcn, aZ8.1795(2) A, bZ
12.9172(4) A, cZ24.9485(9) A, ZZ4, VZ2635.96(14) A ,
3
˚
˚
˚
3
˚
DcZ1.437 g/cm , Mo Ka radiation (lZ0.71073 A), mZ
4. Experimental
4.1. General procedures
0.729 mmK1, crystal dimensions 0.25!0.15!0.05 mm3,
TZ110(2) K. From 12136 reflections, 3014 were unique
(RintZ0.0771). 1956 with IO2s(I) were used in refinement.
Data parameters ratio 3014/170, RZ0.0404, RuZ0.0761.
All reactions and workup procedures were performed under

J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
9765
4.3. Catalytic reactions procedures
2.40 (t, 4H, JZ4.6 Hz); 13C NMR d 136.8, 133.4, 128.6,
127.6, 126.4, 126.1, 67.0. N-1-Phenylallylmorpholine
was not isolated in pure form. Partial H NMR spectra
1
Cross-coupling and Heck reactions were carried out
following reported procedures.3,4 In a typical procedure
for allylic amination, with a ratio substrate/catalystZ
10,000, the palladium/ferrocenyl furylphosphine catalyst
was prepared by stirring a mixture of the diphosphine 1
(0.04 mmol, 22.8 mg) with [Pd(h-C3H5)Cl]2 (0.01 mmol,
3.65 mg) in 10 mL toluene, under argon for 30 min. The
same batch of catalyst was used for more than a week
without any activity decrease (stable in solution in the fridge
under argon). One millilitre of the previously prepared
catalyst solution (thus, 0.001 mmol Pd) was added to the
mixture of allyl acetate and amine (20 mmol) in solvent.
The mixture was stirred at fixed temperature for 1–20 h.
Pure products are obtained after addition of water,
extraction with organic solvents (ether, dichloromethane),
separation, drying of the organic phase (MgSO4), concen-
tration, chromatography on silica gel and possibly distilla-
tion for oily compounds.
was obtained from the mixture: d 5.15 (d, 1H, JZ
17.2 Hz), 5.01 (d, 1H, JZ9.9 Hz).
1
4.4.8. N-3-Phenylallylpyrrolidine. H NMR d 7.30–7.10
(m, 5H), 6.50 (d, 1H, JZ16.0 Hz), 6.20 (dt, 1H, JZ16.0,
6.5 Hz), 3.20 (d, 2H, JZ6.5 Hz), 2.50 (t, 4H, JZ6.5 Hz),
1.80 (m, 4H); 13C NMR d 136.8, 133.4, 128.6, 127.6,
126.4, 126.1, 67.0. N-1-Phenylallylpyrrolidine was not
isolated in pure form. Partial 1H NMR spectra was
obtained from the mixture: d 5.10 (d, 1H, JZ17.4 Hz),
5.05 (d, 1H, JZ10.0 Hz).
1
4.4.9. E-1-(2-Hexenyl)diethylamine. H NMR d 5.59 (dt,
1H, JZ15.4, 6.3 Hz), 5.43 (dt, 1H, JZ15.4, 6.6 Hz), 3.11
(d, 2H, JZ6.3 Hz), 2.58 (q, 4H, JZ7.1 Hz), 1.96 (td, 2H,
JZ7.3, 6.6 Hz), 1.33 (tq, 2H, JZ7.3, 7.3 Hz), 1.02 (t, 6H,
JZ7.1 Hz), 1.33 (t, 3H, JZ7.3 Hz).
4.4. 1H NMR characterization of some coupling products
1
4.4.10. E-1-(2-Hexenyl)morpholine. H NMR d 5.47 (dt,
1H, JZ15.4, 6.3 Hz), 5.35 (dt, 1H, JZ15.4, 6.0 Hz), 3.60 (t,
4H, JZ4.4 Hz), 2.80 (d, 2H, JZ6.3 Hz), 2.31 (d, 4H,
JZ4.4 Hz), 1.90 (td, 2H, JZ7.3, 6.3 Hz), 1.28 (tq, 2H,
JZ7.3, 7.3 Hz), 0.77 (t, 3H, JZ7.3 Hz). 3-(1-Hexenyl)
morpholine was not isolated in pure form. Partial 1H
NMR spectra was obtained from the mixture: d 5.15 (dd,
1H, JZ17.2, 1.5 Hz), 5.05 (dd, 1H, JZ10.0, 1.5 Hz).
References for some of the cross-coupled products are
available.4,36
4.4.1. 4-Acetyl-1,10-biphenyl. 1H NMR (CDCl3): d 8.0
(d, JZ8.5 Hz, 2H), 7.67 (d, JZ8.5 Hz, 2H), 7.61 (d,
JZ7.0 Hz, 2H), 7.45 (dd, JZ7.5, 7.0 Hz, 2H), 7.37 (t,
JZ7.5 Hz, 1H), 2.6 (s, 3H). CAS: 92-91-1.
1
4.4.11. E-1-(2-Hexenyl)pyrrolidine. H NMR d 5.57 (dt,
4.4.2. 4-Methoxy-1,10-biphenyl. 1H NMR (CDCl3): d 7.54
(d, JZ7.4 Hz, 2H), 7.52 (d, JZ8.6 Hz, 2H), 7.40 (dd,
JZ7.4, 7.2 Hz, 2H), 7.29 (t, JZ7.4 Hz, 1H), 6.96 (d,
JZ8.6 Hz, 2H), 3.82 (s, 3H). CAS: 613-37-6.
1H, JZ15.4, 5.8 Hz), 5.45 (dt, 1H, JZ15.4, 6.1 Hz), 3.10
(m, 6H), 2.60 (m, 4H), 1.90 (td, 2H, JZ7.3, 6.1 Hz), 1.28
(tq, 2H, JZ7.3, 7.3 Hz), 0.80 (t, 3H, JZ7.3 Hz). 3-(1-
Hexenyl)pyrrolidine was not isolated in pure form. Partial
1H NMR spectra was obtained from the mixture: d 5.15 (dd,
1H, JZ17.2, 1.5 Hz), 5.05 (dd, 1H, JZ10.0, 1.5 Hz).
4.4.3. E-Butyl 4-acetylcinnamate (coupling anisole/butyl
1
acrylate). H NMR (CDCl3): d 7.55 (d, JZ16.0 Hz, 1H,
ArCH]), 7.37 (d, JZ8.7 Hz, 2H, Ar), 6.80 (d, JZ8.7 Hz,
2H, Ar), 6.21 (d, JZ16.0 Hz, 1H, ]CHCO2Bu), 4.12 (t,
JZ6.6 Hz, 2H, CO2CH2CH2), 3.69 (s, 3H, OMe), 1.60 (tt,
JZ6.8, 6.6 Hz, 2H, CO2CH2CH2), 1.33 (qt, JZ7.3, 6.8 Hz,
2H, CH2CH3), 0.87 (q, JZ7.3 Hz, 3H, CH2CH3). CAS:
173464-57-8.
References and notes
1. (a) Farina, V. Adv. Synth. Catal. 2004, 346, 1519–1521. (b)
Fairlamb, I. J. S. Annu. Rep. Prog. Chem., Sect. B 2004, 100,
113–148.
4.4.4. 4-(2-Phenylethynyl)acetophenone. 1H NMR
(300 MHz, CDCl3): d 7.92 (d, JZ8.1 Hz, 2H), 7.59 (d,
JZ8.1 Hz, 2H), 7.53 (m, 2H), 7.35 (m, 3H), 2.58 (s, 3H).
CAS: 1942-31-0.
2. Farina, V. Adv. Synth. Catal. 2004, 346, 1553–1582.
3. Hierso, J.-C.; Fihri, A.; Amardeil, R.; Meunier, P.; Doucet, H.;
Santelli, M.; Donnadieu, B. Organometallics 2003, 22,
4490–4499.
4.4.5. (4-Methoxyphenyl)phenylacetylene. 1H NMR
(CDCl3): d 7.50–7.48 (m, 2H), 7.45 (d, JZ8.8 Hz, 2H), 7.32
(m, 3H), 6.86 (d, JZ8.8 Hz, 2H), 3.80 (s, 3H). CAS: 7380-
78-1.
4. Hierso, J.-C.; Fihri, A.; Amardeil, R.; Meunier, P.; Doucet, H.;
Santelli, M.; Ivanov, V. V. Org. Lett. 2004, 6, 3473–3476.
5. Hierso, J.-C.; Ivanov, V. V.; Amardeil, R.; Richard, P.;
Meunier, P. Chem. Lett. 2004, 33, 1296–1297.
6. Hierso, J.-C.; Fihri, A.; Ivanov, V. V.; Hanquet, B.; Pirio, N.;
Donnadieu, B.; Rebiere, B.; Amardeil, R.; Meunier, P. J. Am.
Chem. Soc. 2004, 126, 11077–11087.
1
4.4.6. Allyldioctylamine. H NMR d 5.78 (ddt, 1H, JZ
16.9, 10.1, 6.6 Hz), 5.06 (d, 1H, JZ16.9 Hz), 5.00 (d,
1H, JZ10.1 Hz), 3.00 (d, 2H, JZ6.6 Hz), 2.5 (t, 4H, JZ
6.8 Hz), 1.15–1.43 (m, 24H), 0.8 (t, 6H, JZ6.3 Hz).
7. Old, D. W.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc.
1998, 120, 9722–9723.
8. Wolfe, J. P.; Buchwald, S. L. Angew. Chem., Int. Ed. 1999, 38,
2413–2416.
1
4.4.7. N-3-Phenylallylmorpholine. H NMR d 7.35–7.15
(m, 5H), 6.45 (d, 1H, JZ16.4 Hz), 6.18 (dt, 1H, JZ16.4,
6.8 Hz), 3.65 (t, 4H, JZ4.6 Hz), 3.08 (d, 2H, JZ6.8 Hz),
9. Walker, S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L.
Angew. Chem., Int. Ed. 2004, 43, 1871–1876.

9766
J.-C. Hierso et al. / Tetrahedron 61 (2005) 9759–9766
10. Littke, A. F.; Fu, G. C. J. Org. Chem. 1999, 64, 10–11.
11. Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 1999, 37,
3387–3388.
Organic Synthesis; Wiley: Chichester, 1995; p 292: the author
indicates the necessity of bases for alkylation of allyl acetates.
38. Using primary amines monoallylation products and diallyl-
ation products are obtained.
12. Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000, 122,
4020–4028.
39. Albisson, D. A.; Bedford, R. B.; Noelle Scully, P.; Lawrence,
S. E. Chem. Commun. 1998, 2095–2096.
13. Littke, A. F.; Fu, G. C. J. Am. Chem. Soc. 2001, 123,
6989–7000.
40. Albisson, D. A.; Bedford, R. B.; Scully, P. N. Tetrahedron
Lett. 1998, 39, 9793–9796.
14. Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120,
7369–7370.
41. 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, 1073–1076. In this study, kinetics on the Heck reaction
of iodobenzene with styrene have shown that the oxidative
addition step is not rate-limiting and consequently electron-
withdrawing phosphites and phosphorus amidite are more active
catalysts than triarylphosphines.
15. Zapf, A.; Ehrentraut, A.; Beller, M. Angew. Chem., Int. Ed.
2000, 39, 4153–4155.
16. Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113,
9585–9595.
17. Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal
Chemistry; University Science B.: Mill Valley, 1987.
18. Roy, A. H.; Hartwig, J. F. Organometallics 2004, 23,
1533–1541.
42. Casado, A. L.; Espinet, P. J. Am. Chem. Soc. 1998, 120,
8978–8985.
43. Casado, A. L.; Espinet, P.; Gallego, A. M.; Martinez-Ilarduya,
J. M. Chem. Commun. 2001, 339–340.
19. Galardon, E.; Ramdeehul, S.; Brown, J. M.; Cowley, A.; Hii,
K. K.; Jutand, A. Angew. Chem., Int. Ed. 2002, 41, 1760–1763.
20. Kranenburg, M.; Kamer, P. C. J.; Van Leeuwen, P. W. N. M.
Eur. J. Inorg. Chem. 1998, 155–157.
44. Espinet, P.; Echavarren, A. M. Angew. Chem., Int. Ed. 2004,
43, 4704–4734.
45. For a review very clear on this topic see: Bedford, R. B.; Cazin,
C. S. J.; Holder, D. Coord. Chem. Rev. 2004, 248, 2283–2321.
46. For a review very clear on this topic see: Littke, A. F.; Fu, G.
C. Angew. Chem., Int. Ed. 2002, 41, 4176–4211.
47. Even if optimized conditions can be found upon systematic
efforts on variation of the tandem base/solvent, the ratio of
ligand, or the use of additives (see Ref. 22) the coupling
reaction remains somewhat sluggish and the comparison with
strong s-donating ligands is, up to now, unfavourable.
48. Bedford, R. B.; Cazin, C. S. J.; Hazelwood, S. L. Angew.
Chem., Int. Ed. 2002, 41, 4120–4122.
21. Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2001, 123,
1232–1233.
22. Zapf, A.; Beller, M. Chem. Eur. J. 2000, 6, 1830–1833.
23. Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120,
3694–3703.
24. Togni, A.; Hayashi, T. Ferrocenes: Homogeneous Catalysis,
Organic Synthesis, Materials Science; VCH: Weinheim, 1995.
25. Elsagir, A. R.; Gassner, F.; Gorls, H.; Dinjus, E. J. Organomet.
Chem. 2000, 597, 139–145.
26. Kim, T. J.; Kim, Y. H.; Kim, H. S.; Shim, S. C.; Kwak, Y. W.;
Cha, J. S.; Lee, H. S.; Uhm, J. K.; Byun, S. I. Bull. Korean
Chem. Soc. 1992, 13, 588–592.
49. Bedford, R. B.; Hazelwood, S. L.; Limmert, M. E.; Albisson,
D. A.; Draper, S. M.; Scully, P. N.; Coles, S. J.; Hursthouse,
M. B. Chem. Eur. J. 2003, 9, 3216–3227.
27. Andersen, N. G.; Keay, B. A. Chem. Rev. 2001, 101,
997–1030.
50. Under conditions closed to the ones we used: dimethyl-
acetamide, 140–160 8C, 18–20 h, K2CO3 or NaO2CCH3 as a
base.
28. Bandini, A. L.; Banditelli, G.; Cinellu, M. A.; Sanna, G.;
Minghetti, G.; Demartin, F.; Manassero, M. Inorg. Chem.
1989, 28, 404–410.
˚
˚
51. Akermark, B.; Akermark, G.; Hegedus, L. S.; Zetterberg, K. J.
Am. Chem. Soc. 1981, 103, 3037–3040.
29. Bandoli, G.; Dolmella, A. Coord. Chem. Rev. 2000, 209,
161–196.
52. Kuhn, O.; Mayr, H. Angew. Chem., Int. Ed. 1999, 38, 343–346.
See also references therein.
30. Optimized conditions can be found upon systematic efforts on
variation of the tandem base/solvent, the ratio of ligand, or the
use of additives, see Ref. 22.
˚
53. Vitagliano, A.; Akermark, B. J. Organomet. Chem. 1988, 349,
C22–C26.
54. Crociani, B.; Antonaroli, S.; Di Bianca, F.; Canovese, L.;
Visentin, F.; Uguagliati, P. J. Chem. Soc., Dalton Trans. 1994,
1145–1151.
31. Feuerstein, M.; Berthiol, F.; Doucet, H.; Santelli, M. Org.
Biomol. Chem. 2003, 1, 2235–2237.
32. Johannsen, M.; Jorgensen, K. A. Chem. Rev. 1998, 98,
1689–1708.
55. Canovese, L.; Visentin, F.; Uguagliati, P.; Bianca, F. D.;
Antonaroli, S.; Crociani, B. J. Chem. Soc., Dalton Trans. 1994,
3113–3118.
33. Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96,
395–422.
56. Canovese, L.; Visentin, F.; Uguagliati, P.; Crociani, B.; Di
Bianca, F. Inorg. Chim. Acta 1995, 235, 45–50.
57. Canovese, L.; Visentin, F.; Chessa, G.; Niero, A.; Uguagliati,
P. Inorg. Chim. Acta 1999, 293, 44–52.
34. Feuerstein, M.; Laurenti, D.; Doucet, H.; Santelli, M. Chem.
Commun. 2001, 43–44.
35. Feuerstein, M.; Laurenti, D.; Doucet, H.; Santelli, M.
Tetrahedron Lett. 2001, 42, 2313–2315.
58. Tolmachev, A. A.; Ivonin, S. P.; Pinchuk, A. M. Heteroat.
Chem. 1995, 6, 407–412.
36. Feuerstein, M.; Laurenti, D.; Doucet, H.; Santelli, M. J. Mol.
Catal. A: Chem. 2002, 182–183, 471–480 and references
therein.
59. Bishop, J. J.; Davison, A.; Katcher, M. L.; Lichtenberg, D. W.;
Merrill, R. E.; Smart, J. C. J. Organomet. Chem. 1971, 27,
241–249.
37. Tsuji, J. Palladium Reagents and Catalysis: Innovations in