Versatile chemoselectivity in Ni-catalyzed multiple bond carbonylations and cyclocarbonylations in CO2-expanded liquids

Article abstract of DOI:10.1039/b908253k

Selective mono or double carbonylations could be achieved by using CO ₂-expanded liquids in [2 + 2 + 1] carbonylative reactions of alkenes or acetylenes with allyl bromides catalyzed by Ni(i). The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b908253k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Versatile chemoselectivity in Ni-catalyzed multiple bond carbonylations
and cyclocarbonylations in CO2-expanded liquidsw
bc
a
Daniel del Moral,a Anna M. Banet Osuna,ab Alba Cordoba, Josep M. Moreto,
´
´
Jaume Veciana,bc Susagna Ricart*a and Nora Ventosa*bc
Received (in Cambridge, UK) 24th April 2009, Accepted 2nd June 2009
First published as an Advance Article on the web 29th June 2009
DOI: 10.1039/b908253k
Selective mono or double carbonylations could be achieved by using
CO2-expanded liquids in [2 + 2 + 1] carbonylative reactions of
alkenes or acetylenes with allyl bromides catalyzed by Ni(I).
Scheme 2 Catalytic carbonylative cycloaddition of strained alkenes
Multiple C–C bond formation reactions are among the most
important in organic synthesis. In this context, our group has
been working for a long time on the carbonylative cycloaddition
(Pauson–Khand type reactions)1 of allyl halides and alkynes,
and has finally developed a procedure to carry out these
reactions using a catalytic amount of Ni (Scheme 1).2
and allyl halides in acetone.
At present, there is growing interest in the use of compressed
fluids (CF), either in a liquid or supercritical state, as sustainable
solvents in synthetic processes, in particular when applied to
catalytic processes.7 Compressed CO2 is one of the most
attractive alternatives for replacing environmentally-hazardous
conventional organic solvents employed in industrial
reactions, since it is non-toxic, non-flammable and easy to
recover. CO2-expanded liquids are especially attractive for
performing catalytic reactions due to the enhanced solubility
of the reactant gases (e.g. CO), the enhanced transport rates
due to the properties of dense CO2,7,8 the milder operating
pressures (tens of bars) compared to scCO2,9 the resulting
diminished use of organic solvents, and the ability to tune the
solvent power by varying the CO2 composition.10,11
These types of metal-catalyzed [2 + 2 + 1] carbonylative
cycloadditions represent the most straightforward way to synthe-
size the cyclopentanone skeleton, since at least three C–C bonds
are formed in a single experimental operation. This structure is
found in many bioactive natural products of interest (perfumes,
antibiotics, antitumoral agents, etc.).3 Aside from the synthetic
interest, this Ni-catalyzed reaction presents additional advan-
tages; the most promising features being that it gives high yields
(70–95%), takes place at room temperature and atmospheric
pressure, uses a cheap catalyst, is regio and stereoselective
compared with similar processes, and is efficient in the inter-
molecular version.2,4 Also, in early studies, cyclohexenones
resulting from a single carbonylative process were also been
found as a second product.5 However, up until now, by using
liquid solvents, it has not been possible to chemoselectively
enhance the reaction process towards monocarbonylation.
Recently, this procedure has also been extended to strained
olefins (Scheme 2),6 always resulting in a mixture of two
products, corresponding to single and double carbonylation.
Here, we report the outstanding chemoselectivity reached in
this Ni-catalyzed cyclocarbonylation reaction with both alkynes
and strained alkenes by tuning the operational parameters,
using CO2-expanded acetone as the reaction medium instead
of liquid acetone.
An ad hoc experimental set up (Scheme 3) was designed and
built in order to perform the high-pressure catalytic reactions.
The phase behaviour of the reactants was studied on a phase
analyzer based on a variable volume cell with sapphire windows,
which allows visual inspection of the number of phases
present.12 Preliminary solubility experiments were performed
on the different components (reagents and catalyst) involved
in the cyclocarbonylation reaction, both in pure CO2 and
CO2-expanded acetone, at 298 K and pressures up to
Scheme 1 Catalytic carbonylative cycloaddition of alkynes and allyl
halides in acetone.
a
`
Homogeneous Catalysis Group, Institut de Ciencia de Materials de
Barcelona (CSIC), UAB Campus, 08193-Bellaterra, Spain.
E-mail: ricart@icmab.es; Fax: +34 935 805 729;
Tel: +34 935 801 853
b Molecular Nanoscience and Organic Materials Department, Institut
`
de Ciencia de Materials de Barcelona (CSIC), UAB Campus,
08193-Bellaterra, Spain. E-mail: ventosa@icmab.es;
Fax: +34 935 805 729; Tel: +34 935 801 853
Scheme 3 A schematic representation of the equipment used for the
cyclocarbonylation reactions at high-pressure. (P1) High-pressure CO2
pump, (R) chemical reactor, (C) cylinder, (P2) high-pressure liquid
pump, (PI) pressure indicator, (MS) mechanical stirrer, (VL) vent line
and (V) vacuum line.
c Networking Research Center on Bioengineering Biomaterials and
Nanomedicine (CIBER-BBN), 08193-Bellaterra, Spain
w Electronic supplementary information (ESI) available: Detailed expla-
nation of the equipment, experimental procedures and characterization
data of the synthesized products. See DOI: 10.1039/b908253k
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 4723–4725 | 4723

View Article Online
significantly tune the ratio of the two products by changing the
CO content of the reaction medium in the presence of water.
At low CO content, cyclopentanone 3 is obtained with almost
complete selectivity, and the open product is exclusively
produced at higher CO concentrations. Indeed, at high
concentrations, CO efficiently competes with the distal double
bond for the Ni coordination site, precluding its insertion onto
the acyl carbon center, favouring the formation of open
product 4, corresponding to a single carbonylation (Scheme 5).
In view of the good results obtained in the reaction between
norbornenes and allyl halides, we decided to try reactions
between alkynes and allyl halides in CO2-expanded acetone
using the same equipment to see the effect of the CO content
on these reactions. We chose phenylacetylene (5) and 2 as the
model reagents (Scheme 6). To start with, we studied the effect
of the CO content in the absence of CO2 when the reaction was
performed in acetone. In all cases, cyclopentanone 6 was
produced, but we did not detect cyclohexenone 7, corresponding
to the single carbonylation, and a mixture of polyinsertion
products were produced as by-products. The best result with
regard to cyclopentenone formation and reaction rate was at
Scheme 4 The catalytic carbonylative cycloaddition between 1 and 2
in CO2-expanded acetone.
100 bar. The results showed that the tested allyl halides,
norbornenes and alkynes were highly soluble in both pure
CO2 and CO2-expanded acetone. The catalyst, however, was
only soluble in CO2-expanded acetone under certain conditions.
Thus, the reaction was feasible under homogeneous conditions
in the new medium.
We used norbornene (1) and allyl bromide (2) as model
reagents of the reaction between strained alkenes and allyl
halides (Scheme 4).z As mentioned before, using acetone as the
solvent, this reaction always gave a mixture of the mono- and
bicarbonylation products (3 and 4). Under those conditions,
we were only slightly able to tune the ratio between 3 and 4 by
using water as an additive;6 the highest selectivity achieved
being a 68% yield of cyclopentanone 3 and a 20% yield of the
open product 4 (selectivity ratio 77 : 23) in the complete
absence of water.
xCO = 0.0007 (82% yield of cyclopentenone 6). The higher the
CO content, the lower the yield and the slower the rate of the
reaction. As we expected, considering the results obtained with
norbornenes, at xCO = 0.019, the yield of the bicarbonylation
product decreased to 37%, and at xCO = 0.033, a mixture of
unidentified products was obtained.
In Table 1 are summarized the results achieved when
performing the reaction in CO2-expanded acetone at different
operational conditions and in presence or absence of water.
The result in Table 1 entry 2 shows that the cyclocarbonylation
reaction between 1 and the allyl halide gives better yields (both
total yield and yield of the cyclopentanone 3) in CO2-expanded
acetone than in the previously described work (in acetone at
atmospheric pressure (Table 1 entry 1)). Furthermore, contrary
to the results obtained at atmospheric pressure, the use of a
certain amount of H2O did not inhibit the two carbonylation
reactions, but even increased the selectivity in favour of cyclo-
pentanone 3, as seen in Table 1 entry 3. A possible explanation
may arise from the presence of the large amounts of CO2
reacting with H2O, resulting in acidification of the medium.
The decrease of the OHꢁ concentration would then prevent, to
a certain extent, hydroxyl attack at the acyl metal inter-
mediate, which would interrupt the cyclisation. This hypothesis
is currently being studied. In Table 1 entries 3–6, we can see
that increasing the CO content in the liquid phase increases the
ratio of open product to cyclopentanone, with open product 4
being the main product at xCO = 0.0066 (Table 1 entry 5) and
the sole product at xCO = 0.0093 (Table 1 entry 6), despite the
presence of water.
In agreement with the results achieved in acetone, when we
attempted the reaction in CO2-expanded acetone (see Table 2),
the yield of cyclopentenone 6 decreased as the CO content
increased. However, in this solvent medium, we were surprised
to find increasing amounts of cyclohexenone 7, resulting from
a 6-exo-trig cyclization, as xCO augments. When the CO
content in the reaction medium was 0.0043 (Table 2 entry 5),
this product was obtained preferentially as the main product
Scheme 5 The selective formation of products 3 and 4, controlled by
the content of CO in the liquid phase.
We can conclude that by using CO2-expanded acetone in the
catalytic carbonylative cycloaddition between 1 and 2, we can
Scheme 6 Catalytic carbonylative cycloaddition between 5 and 2,
yielding cyclopentenone 6 and cyclohexenone 7.
Table 1 The effect of the CO and H2O content on the cyclocarbonylation reaction of 1 and 2 performed in CO2-expanded acetonea
d
d
e
e
Entry
Ptotal/bar
zCO
zCO
xCO
xCO
Yield (%)
Product ratio 3 : 4
2
2
1b
2b
3c
4c
5c
6c
1.0
—
0.34
—
0.0005
0.0027
0.0027
0.0035
0.0066
0.0093
88
90
92
82
88
71
77 : 23
80 : 20
98 : 2
93 : 7
9 : 91
0 : 100
48.5
48.5
49.0
51.0
53.0
0.92
0.92
0.91
0.88
0.85
0.029
0.029
0.039
0.075
0.108
0.79
0.79
0.78
0.76
0.74
a
b
All reactions were carried out at 25 1C, and the reaction mixture was stirred for 15 h in order to ensure a complete process. In the absence of
H2O. With 1 equivalent of H2O. zi = overall mole fraction of component i. xi = mole fraction of component i in the liquid phase, where the
reaction takes place.
c
d
e
ꢀc
This journal is The Royal Society of Chemistry 2009
4724 | Chem. Commun., 2009, 4723–4725

View Article Online
Table 2 Influence of the CO2 pressure on the carbonylative cycloaddition between 5 and 2, performed in CO2-expanded acetonea
b
b
c
c
Entry
Ptotal/bar
zCO
zCO
xCO
xCO
Yield (%)
Product ratio 6 : 7
2
2
1
2
3
4
5
14.0
31.0
48.0
51.0
59.0
0.79
0.89
0.93
0.93
0.94
0.060
0.030
0.020
0.019
0.016
0.24
0.53
0.79
0.83
0.92
0.0008
0.0011
0.0019
0.0022
0.0043
80
74
74
73
60
100:0
100:0
81:19
73:27
32:68
a
b
All reactions were carried out at 25 1C, and the reaction mixture was stirred for 15 h in order to ensure a complete process. zi = overall mole
fraction of component i. xi = mole fraction of component i in the liquid phase, where the reaction takes place.
c
The CO2 and CO contents in the liquid phase, where the reaction
of the reaction over the cyclopentenone. Although this type of
takes place, were estimated by flash separation calculations using the
ring closure has been seen previously, these products were
Peng Robinson EoS (PR EoS), assuming a ternary CO–CO2–acetone
system and neglecting the effect of the reactants and water. The values
of the binary interaction parameters employed were reported by
always synthesized from internally-substituted allyl halides,
and were obtained as aromatic adducts.5
Lo
´
pez-Castillo et al.,13 who previously reliably used the PR EoS to
Therefore, we can conclude that in the case of acetylenes, in
agreement with the results obtained for strained alkenes, the
formation of cyclohexenone 7 (monocarbonylated adduct) is
favored with increasing CO molar fraction in the liquid phase.
In conclusion, we have reported studies of a highly efficient
catalytic method to selectively synthesize cyclopentanes, cyclo-
hexanes or plainly-carbonylated adducts by tuning the
reaction conditions, namely the water and CO content. The
reaction is intermolecular, starting from products as simple as
allyl halides, alkynes, strained alkenes and CO under mild
reaction conditions by means of a stoichiometric amount of
iron, a catalytic amount of Ni halide and CO2-expanded
acetone as the solvent. This method allows, by changing the
CO content of the reaction medium, the ratio of the products
obtained to be tuned, especially in the case of 1, where we
obtained two different products with almost complete chemo-
selectivity. Moreover, the use of CO2-expanded acetone allowed
a reduction of the acetone consumption by more than 70%.
The reported results open up a wide range of possibilities for
synthesizing more complex cyclopentanones in high yields and
with total chemoselectivity, work that is now in progress.
model CO–CO2-expanded solvent systems. The software used for the
calculations was Hysys Plant.
1 For Pauson–Khand reactions, see: (a) N. Jeong, S. H. Hwang,
Y. W. Lee and J. S. Lim, J. Am. Chem. Soc., 1997, 119, 10549;
(b) O. Geis and H. G. Schmalz, Angew. Chem., Int. Ed., 1998, 37,
911; (c) S. E. Gibson and A. Stevenazzi, Angew. Chem., Int. Ed.,
2003, 42, 1800. For Ni-catalyzed Pauson–Khand-type reactions,
see: (d) M. Zhang and S. L. Buchwald, J. Org. Chem., 1996, 61,
4498.
2 M. Lluısa Nadal, J. Bosch, J. M. Vila, G. Klein, S. Ricart and
¨
J. M. Moreto
3 P. Kraft, J. A. Bajgrowicz, C. Denis and G. Fra
Int. Ed., 2000, 39, 2980.
4 (a) G. Garcıa-Gomez and J. M. Moreto
121, 878; (b) G. Garcıa-Go
2001, 7, 1503; (c) G. Garcıa-Go
Chem., 2001, 1359.
5 (a) F. Camps, J. Coll, J. M. Moreto
1989, 54, 1969; (b) F. Camps, J. M. Moreto
Tetrahedron, 1992, 48, 3147; (c) L. Pages, A. Llebarı
E. Molins, C. Miravitlles and J. M. Moreto, J. Am. Chem. Soc.,
1992, 114, 10449; (d) A. Llebarı and J. M. Moreto
J. Organometal. Chem., 1993, 451, 1.
6 D. del Moral, J. M. Moreto, E. Molins and S. Ricart, Tetrahedron
Lett., 2008, 49, 6947.
´
, J. Am. Chem. Soc., 2005, 127, 10476.
´
ter, Angew. Chem.,
´
´
´
, J. Am. Chem. Soc., 1999,
´
´
mez and J. M. Moreto
´
, Chem.–Eur. J.,
, Eur. J. Org.
´
´
mez and J. M. Moreto
´
´
and J. Torras, J. Org. Chem.,
and L. Pages,
a, F. Camps,
´
´
´
´
a
´
,
´
7 (a) P. G. Jessop, T. Ikariya and R. Noyori, Chem. Rev., 1999, 99,
475; (b) R. S. Oakes, A. A. Clifford and C. M. Rayner, J. Chem.
Soc., Perkin Trans. 1, 2001, 917; (c) R. A. Sheldon, Green Chem.,
2005, 7, 267.
8 (a) D. Prapajati and M. Gohain, Tetrahedron, 2004, 60, 815;
(b) P. Licence, W. K. Gray, M. Sokolova and M. Poliakoff,
J. Am. Chem. Soc., 2005, 127, 293.
9 (a) P. G. Jessop and B. Subramaniam, Chem. Rev., 2007, 107, 2666;
(b) B. Subramaniam, Chem. Eng. Trans., 2002, 2, 857;
(c) B. Subramaniam, C. J. Lyon and V. Arunajatesan, Appl. Catal.,
B, 2002, 37, 279; (d) G. Musie, M. Wei, B. Subramaniam and
D. H. Busch, Coord. Chem. Rev., 2001, 219–221, 789; (e) D. Guha,
H. Jin, M. P. Dudukovic, P. A. Ramachandran and
B. Subramaniam, Chem. Eng. Sci., 2007, 62, 4967.
10 For hydroformylation reactions in SCFs, see: (a) T. Ikariya,
Y. Kayaki, Y. Kishimoto and Y. Noguchi, Prog. Nucl. Energy,
2000, 37, 429; (b) R. Song, J. Zeng and B. Zhong, Catal. Lett.,
2002, 82, 89; (c) Y. Kayaki, Y. Noguchi, S. Iwasa, T. Ikariya and
R. Noyori, Chem. Commun., 1999, 1235; (d) Y. Kishimoto and
T. Ikariya, J. Org. Chem., 2000, 65, 7656; (e) L. Jia, H. Jiang
and J. Li, Green Chem., 1999, 1, 91.
Notes and references
z In a typical experiment, the reactor was manually charged with
NiBr2 (0.43 mmol), NaI (2.55 mmol), Fe powder (8.5 mmol, reducing
agent) and granulated Fe (10.8 mmol, for efficient mixing), and then
sealed. The air inside the reactor and the cylinder was vented through a
vacuum line, and afterwards acetone (1.5 ml) was introduced into the
reactor. CO and CO2 were mixed in a gas cylinder under different
conditions (different partial pressures) for each reaction, and then a
valve connecting the cylinder and the reactor was opened to allow
equilibrium. This valve was kept open during all reactions, as well as
mechanical stirring taking place. After 30 min of reduction, 8.5 mmol
of each substrate pair (1 + 2 or 5 + 2) was slowly added by means of a
high-pressure pump for 3 h. After this, the reaction was left overnight
(approx. 15 h). The following day, the stirring was stopped, the gases
inside the reactor vented into atmosphere and the products collected
with acetone. The solvent was removed and the contents of the flask
transferred to a separation funnel by washing the flask (with the
remaining iron) with dichloromethane. The dark reaction mixture
was treated with portions of 5N HCl solution until no further
discoloration was observed. After washing the organic phase with
water to neutralize it, the organic layer was treated with a solution of
Na2S2O3 (to remove any I2 produced by the oxidation), washed again
with water and dried over MgSO4. The solvent was then removed in a
rotatory evaporator. After this work up, the reaction products were
separated by flash chromatography, and identified and quantified by
1H and 1C NMR spectroscopy. Characterization data for the isolated
products are provided in the ESI.w
11 N. Jeong, S. H. Hwang, Y. W. Lee and J. S. Lim, J. Am. Chem.
Soc., 1997, 119, 10549.
12 (a) M. Munto
Fluids, 2008, 47, 147; (b) M. Munto
J. Supercrit. Fluids, 2008, 47, 290.
13 Z. K. Lopez-Castillo, S. N. Aki, M. A. Stadtherr and
J. F. Brennecke, Ind. Eng. Chem. Res., 2006, 45, 535.
´
, S. Sala, N. Ventosa and J. Veciana, J. Supercrit.
´
, N. Ventosa and J. Veciana,
´
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 4723–4725 | 4725