Palladium catalyzed carboxylation of allylstannanes and boranes using CO2

Article abstract of DOI:10.1039/c0cc03191g

A family of well-defined (η³-allyl)Pd(L)(carboxylate) (L = PR₃ or NHC) complexes are by far the most efficient catalysts reported to date for the catalytic carboxylation of allylstannanes into allylcarboxylates using CO₂. The substrate scope of this reaction is extended to both substituted allylstannanes and allylboranes.

Full text of DOI:10.1039/c0cc03191g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Palladium catalyzed carboxylation of allylstannanes and boranes
using CO2w
Jianguo Wu and Nilay Hazari*
Received 11th August 2010, Accepted 27th October 2010
DOI: 10.1039/c0cc03191g
A family of well-defined (g3-allyl)Pd(L)(carboxylate) (L = PR3 or
NHC) complexes are by far the most efficient catalysts reported to
date for the catalytic carboxylation of allylstannanes into
allylcarboxylates using CO2. The substrate scope of this reaction
is extended to both substituted allylstannanes and allylboranes.
We have recently demonstrated that a family of palladium
allyl complexes of the type (Z1-2-methylallyl)(Z3-2-
methylallyl)Pd(L) (L = PMe3 (1), PEt3 (2), PPh3 (3) or NHC
(4); NHC = 1,3-bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-
imidazol-2-ylidene) react rapidly with CO2 at low temperature
to form the monodentate carboxylates 5–8 (eqn (2)).19
Mechanistic investigations strongly suggest that the reaction
does not proceed via insertion of CO2 into the Z1-allyl Pd–C
bond but through nucleophilic attack of the Z1-allyl terminal
olefin on electrophilic CO2 followed by a palladium mediated
SN2 reaction.19,20 Here we report that compounds 5–8 are by far
the most active catalysts reported to date for the carboxylation
of allylstannanes with CO2; we also extend the substrate scope
to substituted allylstannanes and report preliminary mechanistic
conclusions. Furthermore we have developed new reactions in
which environmentally safer and more synthetically useful
allylboron species are catalytically carboxylated with CO2.
There is great current interest in the utilization of CO2 as a
chemical feedstock.1–6 CO2 is a readily available, inexpensive and
non-toxic source of carbon which in principle could be used in
the synthesis of both commodity chemicals and complex organic
molecules. In particular reactions which result in the catalytic
formation of C–C bonds from CO2 are of interest. One approach
to this problem involves inserting CO2 into the metal–carbon
bond (M–R) of late transition metals to form metal carboxylates
(M–OC(O)R), which contain a weaker M–O bond and are more
likely to undergo further transformations.7–13 In 1997 Nicholas
and Shi demonstrated the first example of the conversion of
allylstannanes into tin carboxylates using CO2 and catalytic
amounts of Pd(PPh3)4 (eqn (1)).14,15 This is a particularly
important reaction as tin carboxylates are used extensively in
industry as stabilizers for polymers and copolymers made from
vinyl chlorides.16,17 Unfortunately, the Pd(PPh3)4 system
required extremely high pressures of CO2 (33 atm), elevated
temperatures (70 1C) and formed a mixture of products. The
mechanism of this reaction has not been thoroughly investigated
but is proposed to involve the insertion of CO2 into palladium(II)
allyls formed in situ. More recently, Johansson and Wendt were
also able to catalyze the formation of a tin carboxylate from an
allylstannane and CO2 using well defined palladium(II)
trifluoroacetate complexes with pincer ligands.18 The reaction
occurred using only 1 atm of CO2 and produced a single product,
but only a single substrate was tested. The first step in the
catalytic cycle was proposed to be the transformation of the
palladium(II) trifluoroacetate species into a catalytically active
allyl complex, which required elevated temperatures. This
process occurred with low efficiency so that all of the
precatalyst was not converted into the active species.18
ð2Þ
Stoichiometric reactions were initially performed between the
monodentate carboxylates 5–8 and tri-n-butyl(2-methylallyl)
stannane to test if it was possible to regenerate the bis(allyl)
compounds 1–4. No reaction was observed at low temperature
but at room temperature partial conversion of the allylstannane to
a tin carboxylate was observed (eqn (3)). In the case of compounds
5, 7 and 8, the corresponding bis(allyl) species 1, 3 and 4 were not
detected, presumably because they are unstable in solution at
room temperature.19 However, in the case of the PEt3 supported
compound 6, the bis(allyl) palladium species 2, which is stable at
room temperature was observed. The greatest degree of
conversion to the tin carboxylate approximately 50% in one
hour was observed with the NHC supported complex 8. No
further conversion was observed after one hour due to
decomposition of the starting material to unidentified products.
ð1Þ
The Department of Chemistry, Yale University, PO Box 208107,
New Haven, Connecticut, 06520, USA. E-mail: nilay.hazari@yale.edu;
Fax: +1 203 432 6144; Tel: +1 203 432 0885
w Electronic supplementary information (ESI) available: Experimental
details and characterizing data. See DOI: 10.1039/c0cc03191g
ð3Þ
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 1069–1071 1069

View Article Online
These results suggested that catalytic carboxylation of
allylstannanes should be possible and complexes 5–8 are
all effective catalysts for the carboxylation of tri-n-butyl
(2-methylallyl)stannane at room temperature using 1 atm of
CO2 (Table 1). Interestingly, poorer results were obtained
when the (Z1-2-methylallyl)(Z3-2-methylallyl)Pd phosphine
complexes 1–3 were used as catalysts. Similarly, systems in
which bis(2-methylallyl)Pd was treated with the appropriate
free ligand in situ to form 1–4 also gave poor results. It is
unclear why the carboxylate species are better catalysts than
the bis(allyl) complexes but this is probably related to the
inherent instability of 1–4,19 which seemed to partially
decompose in the presence of excess tin reagent before CO2
was added. The best results were obtained with the NHC
supported complex 8, which gave the most rapid rates of
both CO2 insertion and transmetallation in our
stoichiometric studies. In most reactions 5 mol% catalyst
loadings were used, however lower loadings could be utilized
with a decrease in reaction rate. If a fresh batch of the
allylstannane and CO2 were added to the reaction mixture at
the end of the reaction, further catalysis could be repeated
without any loss of efficiency (this was repeated three times)
suggesting that catalyst deactivation is slow. Overall, our
conditions are significantly milder than those required by
other systems (room temperature versus 70 1C),14,18 use lower
catalyst loadings, only generate one product, and give higher
yields than previously reported.
Table 2 Carboxylation of different allylstannanes with CO2 in C6D6
using complex 8 as the catalysta
Entry
Substrate
Time/h
Yield (%)
1
2
3
4
5
6
7
8
9
10
Me3Sn(2-methylallyl)
nBu3Sn(2-methylallyl)
Me3Sn(allyl)
2
21
28
21
21
50
23
95
91
80
85
0
90
89
0
nBu3Sn(allyl)
Ph3Sn(allyl)
nBu3Sn(2-ethylallyl)
nBu3Sn(2-phenylallyl)b
nBu2Sn(allyl)2
96 at 60 1C
96 at 60 1C
69 at 50 1C
Sn(allyl)4
nBu3Sn(3,3-dimethylallyl)
0
0
a
The conditions for the reaction were substrate (0.118 mmol), catalyst
8 (3.8 mg, 0.0059 mmol), 1 atm CO2 in 0.25 mL C6D6 at room
b
temperature (unless otherwise stated). 7.6 mg (0.0118 mmol)
catalyst 8 was used.
product are generated, but these can be separated from the
unsubstituted product by column chromatography. The
reaction with the unsubstituted allyl is slower than with the
methylated species and is surprising because in stoichiometric
reactions complexes of the type (Z1-allyl)(Z3-allyl)Pd(L)
(L = PR3 or NHC) decomposed to palladium(I) dimers
before any reaction was observed with CO2.19 Presumably
under the catalytic conditions the relative concentration of
palladium to CO2 is extremely low and the rate of the
bimolecular decomposition reaction is inhibited compared
with CO2 insertion. The fastest reactions occur when the
steric bulk on the allylstannane is reduced (although the
rates for tri-n-butyl and trimethyl stannanes are similar) and
no reaction was observed with the sterically bulky substrate
Ph3Sn(allyl). The reaction is tolerant to substitution in the two
position of the allyl group, however rates were significantly
slower when phenyl and ethyl substituted species were utilized.
A higher catalyst loading was required in the case of the phenyl
species. No reaction was observed with even more electron
deficient groups (such as methyl acetate) in the 2-position of
the allyl and we believe this is a limitation of our catalyst. In all
successful reactions the tin carboxylates could be separated
from the reaction mixture using column chromatography and
The substrate scope of the reaction was investigated using a
number of different allylstannanes and 8 as the catalyst
(Table 2). Along with the substituted substrate, tri-n-butyl
(2-methylallyl)stannane, the unsubstituted reagent tri-n-
butyl(allyl)stannane can also be carboxylated. In this case
two equivalents (relative to the catalyst) of the methylated
Table 1 Carboxylation of tri-n-butyl(2-methylallyl)stannane with
CO2 in C6D6 using different complexes as catalystsa
1
were characterized using H and 13C NMR spectroscopy, IR
Catalyst
Time/h
Yieldb (%)
spectroscopy and high resolution mass spectrometry.
Treatment of the tin carboxylates with HCl generated the
free carboxylic acids in high yield.
1
2
3
5
6
7
8
30
30
60
30
20
50
13
4c
71
80
54
62
90
54
91
5
No reaction was observed with bis and tetra substituted allyl
reagents (entries 8 and 9) which appear to be too electron
deficient to undergo transmetallation reactions with 8 in
stoichiometric reactions. Similarly, no reaction was observed
with substitution in the three position of the allylstannane
(entry 10). Presumably, the steric bulk on the terminal olefin
prevents nucleophilic attack on CO2 by the palladium allyl.
In general boron reagents are more synthetically accessible
and environmentally friendly than tin reagents. Therefore we
wanted to extend our method for the carboxylation of
allylstannanes to allylboranes. A variety of allylboranes were
prepared using literature methods (see ESIw). Table 3 shows
that 8 is an active catalyst for the carboxylation of a series of
allylboranes at room temperature and 1 atm of CO2. To the
best of our knowledge 8 is the first example of a catalyst that
(2-Methylallyl)2Pd + PMe3 (1 in situ)
(2-Methylallyl)2Pd + PEt3 (2 in situ)
(2-Methylallyl)2Pd + PPh3 (3 in situ)
(2-Methylallyl)2Pd + NHC (4 in situ)
30
4c
20
79
5
91
a
The
conditions
for
the
reaction
were
tri-n-butyl(2-
methylallyl)stannane (40.8 mg, 0.118 mmol), catalyst (0.0059 mmol)
b
and 1 atm CO2 in 0.25 mL C6D6 at room temperature. The yield of
the reaction is the yield by 1H NMR spectroscopy after the given
time.1H NMR spectroscopy indicated that only one product was
formed and the product could be isolated by column
chromatography at the end of the reaction with no significant drop
c
in yield. After 5% conversion in the first 4 h no further conversion
was observed by 1H NMR spectroscopy.
c
1070 Chem. Commun., 2011, 47, 1069–1071
This journal is The Royal Society of Chemistry 2011

View Article Online
Table 3 Carboxylation of different allylboranes with CO2 in C6D6
using complex 8 as the catalysta
presumably because the bimolecular decomposition of 4 to
a palladium(I) dimer requires two molecules of palladium,
so under catalytic conditions is relatively slow. Our
stoichiometric studies suggest that transmetallation is rate
determining as the insertion of CO2 into 4 occurs at ꢀ40 1C,19
while transmetallation only occurs at room temperature.
1
Further support for this hypothesis is provided by H NMR
Entry
Substrate
Time/h
Yield (%)
spectroscopy which indicated that 8 is the only species present
during catalysis.
1
2
3
4
5
6
7
(Pinacol)B(2-methylallyl)
(Pinacol)B(allyl)
52
24
23
46
2
81
78
44
36
0
(iPrO)2B(allyl)
In conclusion, we have developed the most efficient catalyst
reported to date for the coupling of allylstannanes with CO2 and
extended the substrate scope of the reaction. Furthermore, we
have described the first report of the coupling of allylboranes
with CO2. In future work we will look to investigate the
mechanism of transmetallation in more detail and optimize
the ligand set so that the catalyst is tolerant to greater
substitution in both the 2- and 3-positions of the allyl group.
(OCH2CH2CH2O)B(allyl)
(Catechol)B(allyl)
(OCH2CH2O)B(allyl)
(OCH(Me)CH(Me)O)B(allyl)
2
2
0
0
a
The conditions for the reaction were substrates (0.118 mmol),
catalyst 8 (3.8 mg, 0.0059 mmol) and 1 atm CO2 in 0.25 mL C6D6 at
room temperature (unless otherwise stated).
Notes and references
1 W. Leitner, Coord. Chem. Rev., 1996, 153, 257–284.
2 X. Yin and J. R. Moss, Coord. Chem. Rev., 1999, 181, 27–59.
3 M. Aresta and A. Dibenedetto, Dalton Trans., 2007, 2975–2992.
4 D. J. Darensbourg, Chem. Rev., 2007, 107, 2388–2410.
´
5 A. Correa and R. Martın, Angew. Chem., Int. Ed., 2009, 48,
6201–6204.
6 T. Sakakura and K. Kohnoa, Chem. Commun., 2009, 1312–1330.
7 K. Ukai, M. Aoki, J. Takaya and N. Iwasawa, J. Am. Chem. Soc.,
2006, 128, 8706–8707.
8 J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2008, 130,
15254–15255.
Scheme 1
can couple allylboranes with CO2. The choice of the ancillary
substituent on boron is crucial and clearly affects the success of
the reaction. In the unsuccessful reactions the substrate and
catalyst often underwent reaction prior to the introduction of
CO2, which led to decomposition of the catalyst. This is
especially true for highly reactive allylboranes, such as
boronic acids. In general, reactions with allylboranes are
slower than with allylstannanes, and the boron carboxylates
are often unstable. However, in most cases they can be
separated from the reaction mixture either through vacuum
transfer or using a silica plug. Importantly, treatment with HCl
rapidly generates the free carboxylic acid in almost
quantitative yield, which is crucial for synthetic applications.
At this stage we propose a straightforward mechanism for
the catalytic carboxylation of allylstannanes and allylboron
reagents using 8 involving two steps: transmetallation and CO2
insertion (Scheme 1). A similar mechanism was proposed by
Johansson and Wendt.18 Clearly the reaction of CO2 with 4
(formed after transmetallation of 8) is faster than the
decomposition of 4, even at room temperature. This is
9 T. Ohishi, M. Nishiura and Z. Hou, Angew. Chem., Int. Ed., 2008,
47, 5792–5795.
10 C. S. Yeung and V. M. Dong, J. Am. Chem. Soc., 2008, 130,
7826–7827.
11 A. Correa and R. Martın, J. Am. Chem. Soc., 2009, 131,
´
15974–15975.
12 I. I. F. Boogaerts and S. P. Nolan, J. Am. Chem. Soc., 2010, 132,
8858–8859.
13 S. Chakraborty, J. Zhang, J. A. Krause and H. Guan, J. Am. Chem.
Soc., 2010, 132, 8872–8873.
14 M. Shi and K. M. Nicholas, J. Am. Chem. Soc., 1997, 119,
5057–5058.
15 R. J. Franks and K. M. Nicholas, Organometallics, 2000, 19,
1458–1460.
16 Kirk-Othmer Concise Encyclopedia of Chemical Technology, ed.
M. Grayson and D. Eckfroth, Wiley, New York, 1985, vol. 12,
pp. 241–246.
17 B. Jousseaume, Microchim. Acta, 1992, 109, 5–12.
18 R. Johansson and O. F. Wendt, Dalton Trans., 2007, 488–492.
19 J. Wu, J. C. Green, N. Hazari, D. P. Hruszkewycz, C. D. Incarvito
and T. J. Schmeier, Organometallics, 2010, DOI: 10.1021/
om1007456.
20 M. T. Johnson, R. Johansson, M. V. Kondrashov, G. Steyl,
M. S. G. Ahlquist, A. Roodt and O. F. Wendt, Organometallics,
2010, 29, 3521–3529.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 1069–1071 1071