Alkoxy- and amidocarbonylation of functionalised aryl and heteroaryl halides catalysed by a Bedford palladacycle and dppf: A comparison with the primary Pd(II) precursors (PhCN)2PdCl2 and Pd(OAc) 2

Article abstract of DOI:10.1039/b615874a

The versatility of a Bedford-type palladacycle 1, namely [{Pd(-Cl){κ²-P,C-P(OC₆H₂-2,4- ^tBu₂)(OC₆H₃-2,4-tBu ₂)₂}}₂], as a primary Pd source, in combination with the ligand bis-1,1′-(diphenylphosphino)ferrocene (dppf) has been established in carbonylation reactions of aryl and heteroaryl bromides with methanol, piperidine and related nucleophiles. Palladacycle 1 has been compared with other primary Pd sources, e.g. (PhCN)₂PdCl₂ and Pd(OAc)₂. The efficacy of the carbonylation processes appear to be linked to the [Pd] concentration, substrate: catalyst ratio, CO pressure and reaction temperature. In amidocarbonylation, double carbonylation is observed for certain organohalides. In the case of 2,5-dibromopyridine, regioselective amination (Hartwig-Buchwald type) also occurs as a side-reaction. This journal is The Royal Society of Chemistry.

Full text of DOI:10.1039/b615874a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Alkoxy- and amidocarbonylation of functionalised aryl and heteroaryl halides
catalysed by a Bedford palladacycle and dppf: a comparison with the primary
Pd(II) precursors (PhCN)2PdCl2 and Pd(OAc)2
Ian J. S. Fairlamb,*a Stephanie Grant,a Peter McCormackb and John Whittallb
Received 1st November 2006, Accepted 11th December 2006
First published as an Advance Article on the web 16th January 2007
DOI: 10.1039/b615874a
The versatility of a Bedford-type palladacycle 1, namely [{Pd(l-Cl){j2-P,C-P(OC6H2-2,4-tBu2)-
(OC6H3-2,4-tBu2)2}}2], as a primary Pd source, in combination with the ligand
bis-1,1ꢀ-(diphenylphosphino)ferrocene (dppf) has been established in carbonylation reactions of aryl
and heteroaryl bromides with methanol, piperidine and related nucleophiles. Palladacycle 1 has been
compared with other primary Pd sources, e.g. (PhCN)2PdCl2 and Pd(OAc)2. The efficacy of the
carbonylation processes appear to be linked to the [Pd] concentration, substrate : catalyst ratio, CO
pressure and reaction temperature. In amidocarbonylation, double carbonylation is observed for
certain organohalides. In the case of 2,5-dibromopyridine, regioselective amination
(Hartwig–Buchwald type) also occurs as a side-reaction.
In certain circumstances double carbonylation is detected; a key
observation first reported by Kobayashi and Tanaka,9a followed by
Introduction
Yamamoto and co-workers.9b It appears that double carbonylation
arises independently of the formation of mono-carbonylation
product, occuring mainly at elevated temperatures.10 For bidentate
ligands, product selectivity is influenced by the length of the
carbon linker connecting the two diphenylphosphino groups;
increasing linker size leads to higher ratios of double carbonylation
product. In addition to these studies, double carbonylation of
aryl iodides has been reported using highly active mono- and
binuclear palladium complexes.11 Kondo and co-workers have also
demonstrated that the monophosphine, t-Bu3P, promotes selective
double carbonylation of electron-deficient aryl iodides.12
Typically, carbonylation reactions involve use of relatively
high loadings of ligand and Pd (1–5 mol% Pd), high pressures
of CO and elevated temperatures.13 We initiated our studies
on the carbonylation of aryl halides with the principal aim
of improving on current known methods. Inspired by recent
studies on palladacycle 1, namely [{Pd(l-Cl){j2-P,C–P(OC6H2-
2,4-tBu2)(OC6H3-2,4-tBu2)2}}2] (Fig. 1), a precatalyst described
by Bedford and co-workers,14 we anticipated that this would be
an ideal Pd source for carbonylation processes. Our rationale for
studying this stems from the ability of 1, and related complexes,
to slowly release catalytically active Pd0 into a suitable reaction
medium.15
Transition metal-catalysed carbonylation of aryl halides in the
presence of an appropriate nucleophile represents a valuable tool
for the selective introduction of carboxylic functionality into
aromatic derivatives. Pd,1 Co,2 and Ni3 have been widely applied
in aryl halide carbonylation. Depending on the nature of the nu-
cleophilic component the products can be aryl esters, amides and
aldehydes.4 Valuable intermediate and final products increasingly
find application in lead structures for use in therapeutic applica-
tions and biological studies.5 Ever-increasing use of Pd catalysis in
synthesis, particularly in cross-coupling, is evident in academic and
industrial settings. Indeed, Pd-catalysed carbonylation processes
are routinely used synthetic transformations. Pionering studies on
Pd-catalysed carbonylation of aryl, benzyl and alkyl halides were
reported independently by Heck,3a and Stille and Wong.6 A paper
detailing the coupling of aryl chlorides in the presence of n-BuOH
(butoxycarbonylation) was reported by Milstein and co-workers in
1989,7 using an electron-rich phosphine ligand “dippp” [dippp =
bis(diisopropylphosphino)propane]; dramatic ligand effects were
seen in this study. More recently, Beller and co-workers have
focused on the butoxycarbonylation of aryl chlorides employing
modified ferrocenyl ligands.8 The major hurdle to be overcome in
aryl chloride carbonylation lies in the coordination of CO to the
metal centre; when bound to the metal, the p-acceptor properties
of CO significantly reduces the reactivity of the metal towards
oxidative insertion into the C–Cl bond—a potential antagonistic
effect for the use of electron-rich r-donor ligands.4 In addition, the
presence of CO accelerates the formation of insoluble non-active
palladium clusters—a well-known phenomena vide infra.
aDepartment of Chemistry, University of York, Heslington, York, UK YO10
5DD. E-mail: ijsf1@york.ac.uk
bStylacats Ltd., The Heath Technical and Business Park, Runcorn, Cheshire,
UK WA7 4QX
Fig. 1 Bedford’s palladacycle 1.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 859–865 | 859
©

View Article Online
Table 1 A comparison of the Pd source in methoxycarbonylation of 2a
In this report the first methoxy- and amidocarbonylation
reactions of aryl halides mediated by 1 at low Pd loadings
(0.05 mol%) are described.16 Catalytic activity has been correlated
against Pd(OAc)2 and/or PdCl2(PhCN)2 for several benchmark
reactions. Alteration of the ligand has a dramatic effect on product
conversion and selectivity (mono-carbonylation versus double-
carbonylation) in amido-carbonylation.
Pd Source
Conversion (%)
Palladacycle 1
Pd(OAc)2
(PhCN)2PdCl2
47
61b
79b
a Reaction conditions as for Scheme 1, using 3 equiv. K2CO3 as the base.
b Formation of insoluble Pd black is observed.
Results and discussion
Table 2 Effect of catalyst loading of 1 in methoxycarbonylation of 2a
The functionalised benchmark substrate 2-bromoaniline 2 was se-
lected for methoxycarbonylation to give 3 (conversions monitored
by GC/MS) (Scheme 1).
[2]/M L−1 [1]/M L−1 S : C ratio
Conversion (%)
0.1
0.5
1
1 × 10−3
1 × 10−3
1 × 10−3
100 : 1 (1 mol% Pd)
500 : 1 (0.2 mol% Pd)
1000 : 1 (0.1 mol% Pd) 63
47
100
a Reaction conditions as for Scheme 1, using 3 equiv. K2CO3 as the base.
23 bar over a period of 3 h showed conversions of 38% and 19%,
respectively. A further pressure increase to 60 bar resulted in no
detectable conversion.
Scheme 1 Methoxycarbonylation of 2-bromoaniline 2.
The substrate : catalyst (S : C) ratio was evaluated to assess
its affect on conversion. On lowering the catalyst loading of 1
to 0.2 mol% Pd loading (S : C, 500 : 1) consumption of 2 was
recorded. At 0.1 mol% Pd loading (S : C, 1000 : 1), the conversion
was reduced to 63%, a conversion higher than that recorded at a
Pd loading an order of magnitude higher (Table 2).
The catalytic activity of palladacycle 1 was compared with
Pd(OAc)2 and (PhCN)2PdCl2 by employing dppf as a ligand
(dppf = bis-1,1ꢀ-(diphenylphosphino)ferrocene), since it was read-
ily available (Pd : dppf = 1 : 3)—note that dppf is a necessary
ligand for carbonylation using these Pd sources.
Using MeOH as the solvent, no consumption of 2 was observed,
possibly due to the poor solubility of dppf in MeOH. Changing to
a solvent mixture of toluene–MeOH (4 : 1, v/v) with 1 (1 mol%) as
the precatalyst, showed 64% conversion. The precise ratio of Pd :
dppf (1 : 3) appears to be important, and the optimum reaction
temperature is 120 ◦C; notably slower conversion is seen at lower
Pd : L ratios (Pd black precipitate is also evident when lower Pd : L
ratios are used) and lower temperatures. Changing the solvent to
dioxane–MeOH (4 : 1, v/v), under the same conditions, increased
the conversion to 82%.
The effect of the base was evaluated next as it has been reported
that organic amine bases react with 1 affording [1−X]·Et3NH
adducts.15 The mineral bases Cs2CO3 and K2CO3 were compared
in parallel reactions with Et3N. Using 1 it was found that 2 was
completely consumed using these inorganic bases (using 3 equiv.).
On a general note, K2CO3 is cheaper than Cs2CO3, and therefore
more attractive (compare K2CO3, £0.02 g−1 and Cs2CO3, £0.25 g−1;
source: Sigma-Aldrich).
The precatalysts (PhCN)2PdCl2 and Pd(OAc)2 were next in-
vestigated (Table 1). Although these Pd sources exhibited higher
conversions for this reaction, extensive formation of palladium
black was seen at the end of the reaction, standing in constrast
to 1, where little palladium black was observed (Table 1). This
indicated a faster release of palladium, faster agglomeration and
the production of insoluble palladium nanoparticles in the case of
(PhCN)2PdCl2 and Pd(OAc)2.
A recent study into the effects of CO pressure by Beller and co-
workers showed that high conversions are seen at low CO pressures
(<10 bar), “moderate” CO pressures (15–70 bar) resulted in a
decrease in the catalytic activity, while high pressures (>75 bar)
show an increase compared to moderate pressures.17 In our study,
in parallel reactions conducted with 1 (1 mol% Pd) at 10 bar and
The inverse correlation18 in catalytic activity no doubt relates
to clustering/agglomeration of mononuclear palladium species in
solution,19 which are either catalytically inactive, after precipita-
tion, or which release mononuclear species back into solution,
albeit slowly in an unfavourable equilibrium. The optimal Pd
loading for this particular reaction is 0.2 mol% 1 (S : C, 500 :
1). It ought to be borne in mind that it has been shown by Dupont
and co-workers that the size of the clusters may be kept small by the
continuous reaction of the aryl halide at the outer rim of any such
clusters.20 Thus at higher aryl halide concentrations, faster erosion
occurs leading to smaller clusters (shown by TEM experiments).
It was stated by de Fries18a that the Pd : substrate ratio is likely
more important than the global palladium concentration per se. In
other informative studies, Trzeciak and co-workers21 have reported
similar findings in methoxycarbonylation of aryl iodides, under
phosphine-free conditions. More broadly, this unusual inverse
correlation has been observed in Heck22 and Sonogashira23 cross-
coupling processes, both of which employed palladacycles as the
Pd source.
We have further established that nitrogen-based nucleophiles
can be utilised in the amidocarbonylation of 2 mediated by
palladacycle 1. For example, piperidine reacts with 2 to afford
the carbonylated product 4 in 73% yield at a S : C ratio of 2000 : 1;
a slightly better yield (80%) was recorded at a S : C ratio of 1000 :
1 (Scheme 2).
4-Methoxyphenyl bromide 5 was next employed as a benchmark
substrate in both methoxycarbonylation and amidocarbonylation
(Scheme 3). At a S : C ratio of 1000 : 1, amidocarbonylation
of 5 occurred to give 6a, but interestingly, for the first time,
formation of the bis-carbonylated product 6b was observed (entry
1, Table 3). All Pd sources showed complete consumption of 5,
however the ratio of mono- to bis-carbonylated product (6a : 6b)
860 | Dalton Trans., 2007, 859–865
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
illustrate the difference in reactivity of piperidine and methanol
and also the catalytic performance of the palladacycle 1 against
(PhCN)2PdCl2 (Fig. 2). Interestingly, addition of NaBArꢀ4 (Arꢀ =
bis-3,5-trifluoromethylbenzene)25 to both 1 and (PhCN)2PdCl2
in situ reveals a dramatic effect. Whilst, the additive clearly retards
the rate of reaction for 1 (entry 8), it completely inhibits the
carbonylation reactions mediated by (PhCN)2PdCl2 (entry 9).
Scheme 2 Amidocarbonylation of 2.
Fig. 2 CO consumption. Key: (a), Pd = 1, Nu = piperidine; (b), Pd =
1, Nu = MeOH; (c), Pd = (PhCN)2PdCl2, Nu = piperidine; (d), Pd =
(PhCN)2PdCl2, Nu = MeOH.
Scheme 3 Amido- and methoxy-carbonylation of 5.
varied according to the type of Pd source used (compare entries
1–3, Table 3). Curiously, palladacycle 1 exhibits a similar CO
consumption profile to both (PhCN)2PdCl2 and Pd(OAc)2, while
different selectivities24 are observed, e.g. the rates of carbonylation
are similar for all the Pd sources tested. Methoxycarbonylation of 5
gave the mono-carbonylation product 7, exclusively. On lowering
the catalyst loading to 0.05 mol% Pd (S : C 2000 : 1), product
conversion was maintained with similar product selectivity (entry
4). Switching the precatalyst to (PhCN)2PdCl2 resulted in lower
conversion and selectivity (entry 5). It is important to note that
increasing the temperature from 120 ◦C to 140 ◦C for both
palladacycle 1 and (PhCN)2PdCl2 results in an increase in selec-
tivitity towards 6a; for (PhCN)2PdCl2 conversion is dramatically
improved. Use of methanol as the nucleophile at a S : C ratio
of 2000 : 1 affords the mono-carbonylated product 6a exclusively
(entry 6), although modest conversion is recorded (PhCN)2PdCl2
(entry 7). The CO consumption profiles for entries 4–5 simply
Some other selected ligands were evaulated against dppf in
parallel reactions of 5 to 6a/6b to highlight the usefulness of this
cheap and readily available ligand, when used in combination with
1 (Fig. 3). Whilst Xantphos proved as an effective ligand as dppf
in terms of product conversion (88% versus 92%, respectively),
product selectivity was slightly reduced (6a : 6b, 91 : 9 versus
97 : 3, respectively). On the other hand, the P,S-ligands 8a and
8b exhibited conversions of only 10% (6a : 6b, 1 : 1) and 9%
(6a exclusively), respectively. A derivative of the ESPHOS class,
Table 3 A comparison of the Pd source in amidocarbonylation of 5a
Entry Pd Source
S : C
Conversion (%) 6a : 6b
1
2
3
4
Palladacycle 1
(PhCN)2PdCl2
Pd(OAc)2
1000 : 1 >99
1000 : 1 >99
1000 : 1 >99
2000 : 1 >99
>99c
99 : 1
91 : 9
94 : 6
96 : 4
98 : 2
93 : 7
97 : 3
100 : 0
100 : 0
N.D.
—
Palladacycle 1
5
(PhCN)2PdCl2
2000 : 1
40
>99c
6
7
8
9
Palladacycle 1d
2000 : 1 >99
d
(PhCN)2PdCl2
2000 : 1
56
Palladacycle 1 + NaBArꢀ4 2000 : 1 ∼40
(PhCN)2PdCl2 + NaBArꢀ4 2000 : 1
0
a See Scheme 3 for reaction conditions. b Det◦ermined by GC-MS; N.D. =
not determined. c Reaction conducted at 140 C (after 1 h). d Using MeOH
(1.5 equiv.) as the nucleophile. e 1 equiv. of NaBArꢀ4 per Pd.
Fig. 3 Other ligands tested in the amidocarbonylation of 5 with
piperidine.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 859–865 | 861
©

View Article Online
namely 8c, developed by Wills and co-workers,26 also exhibited
low conversion (5%) and poor selectivity (6a : 6b, 1.7 : 1).
The bicyclo[3.2.0]heptanyl diphosphinate ligand, B[3.2.0]-
DPO,27 was also tested in several carbonylation reactions, however
the results were not reproducible.28
We also evaluated the catalyst system 1–dppf against
(PhCN)2PdCl2–dppf, described as a versatile catalyst system for
amidocarbonylation of unprotected 5-bromo-1H-indole 9 with
piperidine by Beller and co-workers (Scheme 4 and Table 4).29
Interest in this substrate stems from its potential application in
the synthesis of ligands that interact with the serotonin (5-HT)
subtype 2A receptor (a CNS target).30
gave an identical result to dppf (entry 4, Table 4; compare entry 2).
Gratifyingly, with palladacycle 1 near quantative conversion for
both dppf and Xantphos ligands was established (entries 5 and 6,
Table 4); the temperature can also be lowered from 130 to 120 ◦C.
Having established an improved protocol for amidocar-
bonylation of 9, regioselective amidocarbonylation of 2,5-
dibromopyridines 11 and 13 with piperidine was tested (Scheme 5,
Table 5).31 In general, carbonylation/coupling is expected at the
2-position (the more electron-deficient carbon centre). Several
products are possible in this reaction, e.g. the expected products
12a and 14a, and double functionalised products 12b and 14b.
Under a range of conditions these products were formed (Table 5).
Scheme 4 Amidocarbonylation of 9.
We reproduced an experiment reported by Beller using PPh3
as a ligand and (PhCN)2PdCl2 as the Pd source, as a benchmark
(entry 1, Table 4)—under identical conditions (reagent quantities
and ratios, temperature, CO pressure etc.) 30% conversion to
compound 10 was recorded, compared to the reported 24%
conversion.29 Use of dppf was described under identical conditions
in this report, except that the reaction time was 20 h. To maintain
a fair comparison the conversion was analysed after 12 h for dppf
(entry 2, Table 4), which showed a 50% conversion (reported29 92%
after 20 h).
Scheme 5 Amidocarbonylation of 2,5-dibromopyridines 11 and 13.
However, for the first time we also detected the formation of the
Hartwig–Buchwald amination32 products 12c and 14c.
Using K2CO3 as the base at 70 ◦C and a S : C ratio of 2000 : 1
2,5-dibromopyridine 11 reacted with piperidine to afford products
12a, 12b and 12c in a rati◦o of 12 : 8 : 1 (entry 1, Table 5). Increasing
the temperature to 100 C dramatically increased the formation
of the amination product 12c (entry 2, Table 5). Under these
conditions, 12a was the minor product. Reducing the temperature
to 50 ◦C served to increase the selectivity for 12a, providing that
the Pd loading was increased to 0.08 mol% (S : C., 1250 : 1) (entry
3, Table 5).
Lowering the Pd loading to 0.1 mol% served only to reduce
product conversion (entry 3, Table 4). Use of Xantphos as a ligand
Table 4 Amidocarbonylation of 5-bromo-1H-indole 9
Increasing the Pd loading to 0.33 mol% (S : C, ∼300 : 1) improves
the conversion to 96%, which moreover increases the selectivity
for 12a. Changing the base to NaOAc or DABCO results in lower
conversions at 50 ◦C (compare entries 5 and 6 with 3, Table 5).
Interestingly, for 2,5-dibromo-3-methylpyridine 13 the forma-
tion of the amination product 14c is less prevalent than was seen
with 11 at 100 ◦C, which affords products 14a, 14b and 14c in
a 15 : 1.5 : 1 ratio (entry 7, Table 5). Lowering the temperature
to 60 ◦C results in a better selectivity for 14a, however lower
Entry Pd Source
Pd/mol% Ligand
T/◦C Conv. (%)
1
2
3
4
5
6
(PhCN)2PdCl2
(PhCN)2PdCl2
(PhCN)2PdCl2 0.1
(PhCN)2PdCl2
Palladacycle 1
Palladacycle 1
1
1
PPh3
dppf
dppf
130
130
130
30
50
9
1
1
1
Xantphos 130
dppf 120
Xantphos 120
50
>95
>95
a See Scheme 4 for reaction conditions.
Table 5 Amidocarbonylation of dibromopyridines 11 and 13
Entry
Pd/mol%
R
T/◦C
Base
Conv. (%)
Product ratio
1
2
3
4
5
6
7
8
0.05
0.05
0.08
0.33
0.08
0.08
0.08
0.08
H
H
H
H
H
H
Me
Me
70
100
50
50
50
50
100
60
K2CO3
K2CO3
K2CO3
K2CO3
NaOAc
DABCO
K2CO3
K2CO3
100
100
75
96
15
21
89
40
12a : 12b : 12c, 12 : 8 : 1
12a : 12b : 12c, 1 : 5 : 12
12a : 12b : 12c, 11 : 1 : 0
12a : 12b : 12c, 30 : 1 : 0
12a : 12b : 12c, 3 : 2 : 0
12a : 12b : 12c, 1 : 1 : 0
14a : 14b : 14c, 15 : 1.5 : 1
14a : 14b : 14c, 20 : 1 : 0
a See Scheme 5 for reaction conditions; piperidine (1.5 equiv.) used as the nucleophile.
862 | Dalton Trans., 2007, 859–865
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
conversion is recorded (entry 8, Table 5), highlighting that 13 is a
more deactivated substrate than 11.
To explore the potential scope of 1 in mediating efficient amido-
carbonylation, a series of other organohalide and amine substrates
were evaluated (Table 6). 4-Methoxyphenyl bromide 5 reacted
with cyclohexylamine 15 to afford mainly the monocarbonylated
product 16a; the bis-carbonylated product 16b was also formed
(entry 1, Table 6). Morpholine 17 was also found to be reactive
towards 5, however an equal mixture of both 18a and 18b were
formed (entry 2, Table 6). The deactivated amine 19 (containing
two electron-withdrawing meta-methoxy substituents) reacted
well with 5, which was highly selective for the monocarbonylation
product 20a (entry 3, Table 6). Piperazines 21 and 23 also reacted
selectively with 5 to afford the monocarbonylation products 22a
and 24a, respectively (entries 4 and 5, Table 6). Against a more
activated aryl halide substrate 25 it was found that piperizine
21 showed a slightly lower conversion and selectivity than for
the reaction with 5 (compare entries 4 and 6, Table 6). 4-
bromoaniline 27, a deactivated substrate which could also compete
as a nucleophile in amidocarbonylation, showed good conversion
but modest selectivity for the monocarbonylation product 28a
(entry 7, Table 6).
Scheme 6 Degradation of 1 in the presence of dppf.
be expected to ameliorate catalyst stability in the carbonylation
reactions.
Clearly for some aryl halide substrates, e.g. 2,5-dibromopyridine
11, a PdII catalyst resting state could equally be as important.33
Indeed, the dramatic differences in product selectivity observed
for this substrate on changing the reaction temperature from 70
to 100 ◦C suggests that there is a higher concentration of the PdII
intermediate at higher temperatures.
Conclusions
In this paper we have established that palladacycle 1, originally
developed by Bedford and co-workers for use in cross-coupling
reactions, is useful for both amidocarbonylation and methoxycar-
bonylation processes. It has been demonstrated that 1 is generally
a superior source of Pd over other primary sources of Pd such
as Pd(OAc)2 and (PhCN)2PdCl2. Although it is evident that
Pd(OAc)2 and (PhCN)2PdCl2 are useful Pd sources, generally
higher temperatures are required for equivalent conversions and
product selectivities seen using 1.
Experimental
General procedure for carbonylation
Carbonylation reactions were carried out at 120 ◦C, unless
otherwise stated, in a stainless steel Parr-Multi channel reactor
with a stirrer rate of 1000 rpm. Carbon monoxide pressure of
9–10 bar was applied to the reactions which were monitored by
following CO consumption. For analytical purposes, samples from
reactions were combined and purified by column chromatography
on silica-gel eluting with ethyl acetate–hexane mixtures.
Based on previous literature precedent, and the results detailed
herein, it can be proposed that the palladacyclic structure slowly
breaks down under the reaction conditions releasing Pd0 into the
catalytic cycle, particularly given that relatively high temperatures
and modest CO pressures were used. The importance of the Pd :
dppf ratio of 1 : 3 suggests that “(dppf)Pd0” is a potential catalyt-
ically active species. However, (dppf)2Pd0 or (dppf)Pd0P(OAr)3
(Ar = C6H3-2,4-t-Bu2) complexes could also be important as
catalyst resting states. Bedford and co-workers14c have previously
shown that the Pt variant of 1 reacts cleanly with dppf to generate
[(dppf)PtIIP(OAr)3]X complexes. Reaction of 1 with 2 equiv. of
HCl was also shown to result in de-orthometallation. It is therefore
not unresonable to suggest that 1 reacts with dppf to give inter-
mediate I, followed by de-orthometallation to give II, which can
then be reduced in situ to III (Scheme 6). The p-acidic phosphite
ligand, as well as CO, would readily coordinate to and stabilise
“(dppf)Pd0” both kinetically and thermodynamically—reversible
decoordination of either ligand would release “(dppf)Pd0” back
into the catalytic cycle. The differences seen between 1, Pd(OAc)2
and (PhCN)2PdCl2, suggest that the phosphite ligand is associated
with the catalytic cycle. In keeping with the proposal that mixed
phosphite–phosphine palladium complexes derived from 1, e.g.
“Pd(P(OAr)3)PCy3”, increase catalyst longevity in Suzuki cross-
couplings of deactivated aryl halides,14c the phosphite ligand would
Using a substrate : catalyst ratio of 2000 : 1. To oven-dried
stainless steel bombs containing a “propeller” magnetic stirrer
bar was added aryl halide (15 mmol, 1 equiv.), K2CO3 (45 mmol,
3 equiv.), palladium source (0.05 mol%), dppf (0.15 mol%). The
stainless steel vessels were evacuated and back-filled with argon
gas (5 to 7 times) and sealed using a rubber septum, before
addition of a solution of dried degassed dioxane (by three freeze–
pump–thaw cycles). A solution containing the desired quantity
of nucleophile in 14 mL of solvent was then added. The bombs
were assembled and pressurised, then depressurised with argon (4
times), followed by CO (5 to 7 times). Reactions were worked-
up by filtration through a silica-gel plug washed with EtOAc
or CH2Cl2, which were analysed directly by GC-MS against
hexadecane (as the internal standard). All products, and ratios
of mono- and bis-carbonylated products, were characterised and
calculated principally by GC-MS. Product purity was determined
by 1H and 13C NMR spectra. Spectroscopic data were in agreement
with those reported (see citations indicated in the text against
specific benchmark reactions).
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 859–865 | 863
©

View Article Online
Table 6 Substrate scope in amidocarbonylation
Entry
1
ArX
Amine
Products
T/◦C
Conv. (%)
93
Product ratio
16a : 16b, 84 : 16
130
2
5
5
120
120
58
18a : 18b, 1 : 1
/>
3b
100
20a : 20b, 98 : 2
4
5
130
92
22a : 22b, 96 : 4
5
6
130
100
100
24a : 24b, 98 : 2
84
26a : 26b, 92 : 8
7
130
64
28a : 28b, 66 : 34
a For general reaction conditions: Pd 1 (0.05 mol%), dppf (0.15 mol%), aryl halide (1 equiv.), amine (1.5 equiv.), K2CO3 (3 equiv.), CO (9 bar), dioxane,
130 ◦C, 14 h; unless otherwise stated the substrate : catalyst ratio = 2000 : 1. b Substrate : catalyst = 1000 : 1.
864 | Dalton Trans., 2007, 859–865
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
17 W. Ma¨gerlein, M. Beller and A. F. Indolese, J. Mol. Catal. A: Chem.,
2000, 156, 213.
Acknowledgements
18 (a) J. G. de Vries, Dalton Trans., 2006, 421; (b) I. J. S. Fairlamb,
R. J. K. Taylor, J. L. Serrano and G. Sanchez, New J. Chem., 2006, 30,
1695.
19 M. T. Reetz and J. G. de Vries, Chem. Commun., 2004, 1559.
20 C. C. Cassol, A. P. Umpierre, G. Machado, S. I. Wolke and J. Dupont,
J. Am. Chem. Soc., 2005, 127, 3298.
21 (a) A. M. Trzeciak, W. Wojtko´w, J. J. Zio´łkowski, J. Wrzyszcz and M.
Zawadzki, New J. Chem., 2004, 28, 859; (b) A. Gniewek, J. J. Zio´łkowski,
A. M. Trzeciak and L. Kepinski, J. Catal., 2006, 239, 272; (c) A.
Gniewek, A. M. Trzeciak, J. J. Zio´łkowski, L. Kepinski, J. Wrzyszcz
and W. Tylus, J. Catal., 2005, 229, 332.
We are grateful to Prof. Stanley M. Roberts (Liverpool University)
for supporting this research and Stylacats Ltd. for funding.
The EPSRC are acknowledged for a Case for New Academics
(CNA) studentship (for S. G.). I. J. S. F. is grateful to the Royal
Society (UK) for a University Research Fellowship and generous
equipment grant.
References and notes
22 A. H. M. de Vries, J. M. C. A. Mulders, J. H. M. Mommers, H. J. W.
Henderickx and J. G. de Vries, Org. Lett., 2003, 5, 3285.
23 I. J. S. Fairlamb, A. R. Kapdi, A. F. Lee, G. Sa´nchez, G. Lo´pez,
J. L. Serrano, L. Garc´ıa, J. Pe´rez and E. Pe´rez, Dalton Trans., 2004,
3970.
1 A. Schoenberg, I. Bartoletti and R. F. Heck, J. Org. Chem., 1974, 39,
3318.
2 (a) K. Kudo, T. Shibata, T. Kashimura, S. Mori and N. Sugita, Chem.
Lett., 1987, 577; (b) M. Foa, F. Francalanci, E. Bencini and A. Gardano,
J. Organomet. Chem., 1973, 51, 381.
24 It is unclear at this stage whether the anionic ligand (e.g. chloride or
acetate) controls the product selectivity or other ligands. It is possible
that the p-accepting phosphite ligand in 1 could play a decisive role in
promoting selectivity for 6a. For discussions of halide and pseudohalide
effects in Pd-mediated reactions, see: I. J. S. Fairlamb, R. J. K. Taylor,
J. L. Serrano and G. Sanchez, New. J. Chem., 2006, 30, 1695, and also
ref. 18b.
25 (a) M. Brookhart, B. Grant and A. F. Volpe, Organometallics, 1992, 11,
3920; (b) N. A. Yakelis and R. G. Bergmann, Organometallics, 2005,
24, 3579.
26 (a) S. Breeden, D. J. Cole-Hamilton, D. F. Foster, G. J. Schwarz and
M. Wills, Angew. Chem., Int. Ed., 2000, 39, 4106; (b) S. W. Breeden
and M. Wills, J. Org. Chem., 1999, 64, 9735; (c) G. J. Clarkson, J. R.
Ansell, D. J. Cole-Hamilton, P. J. Pogorzelec, J. Whittell and M. Wills,
Tetrahedron: Asymmetry, 2004, 15, 1787.
27 (a) N. Derrien, C. B. Dousson, S. M. Roberts, U. Berens, M. Burk and
M. J. Ohff, Tetrahedron: Asymmetry, 1999, 10, 3341; (b) B. Adger,
U. Berens, M. J. Griffiths, M. J. Kelly, R. McCague, J. A. Miller,
C. F. Palmer, S. M. Roberts, R. Selke, U. Vitinus and G. Ward, Chem.
Commun., 1997, 1713; (c) I. J. S. Fairlamb, S. Grant, A. C. Whitwood,
J. Whitthall, A. S. Batsanov and J. C. Collings, J. Organomet. Chem.,
2005, 690, 4462; (d) I. J. S. Fairlamb, S. Grant, S. Tommasi, J. M. Lynam,
M. Bandini, H. Dong, Z. Lin and A. C. Whitwood, Adv. Synth. Catal.,
2006, 348, 2515.
28 It appears that ligand elimination occurs in B[3.2.0]DPO ligands,
which is accelerated by the presence of water and/or trace acid. A
comprehensive study will be reported in due course: S. Grant and
I. J. S. Fairlamb, unpublished results. For a discussion concerning ligand
elimination in phosphinite ligands see: (a) I. Pryjomsk, H. Bartosz-
Bechowski, Z. Ciunik, A. M. Trzeciak and J. J. Zio´łkowski, Dalton
Trans., 2006, 213; (b) A. M. Trzeciak, H. Bartosz-Bechowski, Z. Ciunik,
K. Niesyty and J. J. Zio´łkowski, Can. J. Chem., 2001, 79, 752; and also
ref. 27c.
29 K. Kumar, A. Zapf, D. Michalik, A. Tillack, T. Heinrich, H. Bo¨ttcher,
M. Arlt and M. Beller, Org. Lett., 2004, 6, 7.
30 H. Bo¨ttcher, J. Marz, H. Greiner, J. Harting, G. Bartoszyk, C.
Seyfried and C. V. Amsterdam, Merck Patent GmbH, Germany, WO,
2 001 007 434, 2001.
3 (a) R. F. Heck, J. Organomet. Chem., 1963, 85, 2013; (b) E. J. Corey
and L. S. Hegedus, J. Am. Chem. Soc., 1969, 91, 1231; (c) I. Amer and
H. Alper, J. Org. Chem., 1988, 53, 5147.
4 W. Ma¨gerlein, A. F. Indolese and M. Beller, J. Organomet. Chem., 2002,
641, 30.
5 J. S. Carey, D. Laffan, C. Thomson and M. T. Williams, Org. Biomol.
Chem., 2006, 4, 2337.
6 J. K. Stille and P. K. Wong, J. Org. Chem., 1975, 40, 532.
7 (a) Y. Ben-David, M. Portnoy and D. Milstein, J. Am. Chem. Soc., 1989,
111, 8742; (b) M. Portnoy and D. Milstein, Organometallics, 1993, 12,
1655.
8 K. Kumar, A. Zapf, D. Michalik, A. Tilack, T. Heinrich, H. Bo¨ttcher,
M. Arlt and M. Beller, Org. Lett., 2004, 6, 7.
9 (a) T. Kobayashi and M. Tanaka, J. Organomet. Chem., 1982, 233, C64;
(b) F. Ozawa, H. Soyama, T. Yamamoto and A. Yamamoto, Tetrahedron
Lett., 1982, 23, 3382.
10 (a) F. Ozawa, T. Sugimoto, Y. Yuasa, M. Santra, T. Yamamoto and A.
Yamamoto, Organometallics, 1984, 3, 683; (b) F. Ozawa, T. Sugimoto,
T. Yamamoto and A. Yamamoto, Organometallics, 1984, 3, 692.
11 N. Tsukada, Y. Ohba and Y. T. Inoue, J. Organomet. Chem., 2003, 687,
436.
12 (a) M. Iizuka and Y. Kondo, Chem. Commun., 2006, 1739. For related
reports describing the use of t-Bu3P in carbonylation processes, see:
(b) S. Couve-Bonnaire, J.-F. Carpentier, A. Mortreux and Y. Castanet,
Tetrahedron, 2003, 59, 2793; (c) X. Wu, P. Nilsson and M. Larhed,
J. Org. Chem., 2005, 70, 346; (d) F. Karimi, J. Barletta and B. La˚ngstro¨m,
Eur. J. Org. Chem., 2005, 2374.
13 Y. Bessard and J. P. Roduit, Tetrahedron, 1999, 55, 393.
14 (a) D. A. Albisson, R. B. Bedford, S. E. Lawrence and P. N. Scully,
Chem. Commun., 1998, 2095; (b) R. B. Bedford and S. L. Welch,
Chem. Commun., 2001, 129; (c) R. B. Bedford, S. L. Hazelwood, M. E.
Limmert, D. A. Albisson, S. M. Draper, P. Noelle Scully, S. J. Coles
and M. B. Hursthouse, Chem.–Eur. J., 2003, 9, 3216.
15 R. B. Bedford, S. L. Hazelwood, P. N. Horton and M. B. Hursthouse,
Dalton Trans., 2003, 4164.
16 For carbonylation of aryl iodides mediated by a dimeric palladacycle,
see: (a) C. Ramesh, Y. Kubota, M. Miwa and Y. Sugi, Synthesis,
2002, 2171. For the synthesis of isoindolinones via a three-component
cyclative carbonylation-amination cascade mediated by a palladacycle,
see: (b) R. Grigg, L. X. Zhang, S. Collard and A. Keep, Tetrahedron
Lett., 2003, 44, 6979. For reviews relating to the use of palladacycles
in catalysis and other applications, see: (c) J. Dupont, C. S. Consorti
and J. Spencer, Chem. Rev., 2005, 105, 2527; (d) I. P. Beletskaya and
A. V. Cheprakov, J. Organomet. Chem., 2004, 689, 4055; (e) B. C. G.
Soderberg, Coord. Chem. Rev., 2004, 248, 1085.
31 G. G. Wu, Y. Wong and M. Poirier, Org. Lett., 1999, 1, 745.
32 (a) J. F. Hartwig, Acc. Chem. Res., 1998, 31, 852; (b) J. F. Hartwig,
in Organopalladium Chemistry for Organic Synthesis, ed. E.-i. Negishi,
Wiley-Interscience, New York, 2002, vol. 1, p. 1051; (c) B. H. Yang and
S. L. Buchwald, J. Organomet. Chem., 1999, 576, 125.
33 A. Beeby, S. Bettington, I. J. S. Fairlamb, A. E. Goeta, A. Kapdi and
A. L. Thompson, New. J. Chem., 2004, 28, 600.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 859–865 | 865
©