Baeyer-Villiger oxidation in supercritical CO2 with potassium peroxomonosulfate supported on acidic silica gel

Article abstract of DOI:10.1021/jo060753p

Supercritical carbon dioxide (scCO₂) is an efficient reaction medium to perform the Baeyer-Villiger oxidation with hydrated silica-supported potassium peroxomonosulfate (h-SiO₂·KHSO₅) under flow-through conditions. Hydration modulates the reactivity of the active surface by softening the acidity of the KHSO₄ present in the supported reagent. The reaction in scCO₂ is much more efficient than in n-hexane under similar conditions, which is attributed to better transport and solvating properties of the supercritical medium with regard to n-hexane.

Full text of DOI:10.1021/jo060753p

Baeyer-Villiger Oxidation in Supercritical CO2 with Potassium
Peroxomonosulfate Supported on Acidic Silica Gel
Mar´ıa E. Gonza´lez-Nu´n˜ez, Rossella Mello, Andrea Olmos, and Gregorio Asensio*
Departamento de Qu´ımica Orga´nica, UniVersidad de Valencia, AVda. V. Andre´s Estelle´s s/n,
46100-Burjassot (Valencia), Spain
gregorio.asensio@uV.es
ReceiVed April 10, 2006
Supercritical carbon dioxide (scCO2) is an efficient reaction medium to perform the Baeyer-Villiger
oxidation with hydrated silica-supported potassium peroxomonosulfate (h-SiO2‚KHSO5) under flow-
through conditions. Hydration modulates the reactivity of the active surface by softening the acidity of
the KHSO4 present in the supported reagent. The reaction in scCO2 is much more efficient than in n-hexane
under similar conditions, which is attributed to better transport and solvating properties of the supercritical
medium with regard to n-hexane.
SCHEME 1. Baeyer-Villiger Reaction with SiO2‚KHSO5
Introduction
The Baeyer-Villiger oxidation of ketones 1 into esters 2 is
one of the most widely applied transformations in organic
synthesis.1 However, the standard protocol for this selective
oxidation still suffers several disadvantages. The most common
reagents for performing these transformations are organic
peracids,1 resulting in the formation of 1 equiv of the corre-
sponding carboxylic acid, which has to be separated from the
reaction products. Moreover, organic peracids are expensive and/
or hazardous (because of shock sensitivity), which limits their
commercial application. Also, Baeyer-Villiger oxidation gener-
ally requires the use of organic solvents, typically dichlo-
romethane or aromatics, which further enhances the environ-
mental impact of the process.
We have recently reported2 that potassium peroxomonosulfate
deposited onto silica, SiO2‚KHSO5, efficiently reacts with
ketones in dichloromethane to provide the corresponding esters
or lactones in quantitative yields. Potassium peroxomonosulfate
is a good alternative to organic peracids in the Baeyer-Villiger
reaction (Scheme 1) since it is inexpensive and safe. In addition,
silica-supported potassium peroxomonosulfate avoids the hy-
drolysis of reaction products, an extensive side process associ-
ated with the use of aqueous solutions of Oxone, and offers
advantages in relation to the use of supported reagents in organic
synthesis, such as facilitating the separation of the products from
the reduced reagent. Further improvement of the Baeyer-
Villiger reaction with supported potassium peroxomonosulfate
could be achieved by avoiding the use of chlorinated organic
solvents.
Supercritical carbon dioxide (scCO2), which has readily
accessible critical conditions (Tc ) 31.0 °C, Pc ) 73.8 bar), a
truly benign character, and low cost, is a frequently discussed
alternative reaction medium for chemical synthesis.3 Further-
more, there is an existing technology platform for the use of
scCO2 in large-scale applications in both the food and nutrition
industries. These properties have prompted intense research to
further develop the potential of scCO2 as an alternative solvent
for green chemistry. Transport properties of scCO2 (adjustable
(1) (a) Krow, G. R. Org. React. 1993, 43, 251. (b) Renz, M.; Meunier,
B. Eur. J. Org. Chem. 1999, 737. (c) tenBrink, G.-J.; Arends, I. W. C. E.;
Sheldon, R. A. Chem. ReV. 2004, 104, 4150.
(2) Gonza´lez-Nu´n˜ez, M. E.; Mello, R.; Olmos, A.; Asensio, G. J. Org.
Chem. 2005, 70, 10879-10882.
(3) (a) Leitner, W. Acc. Chem. Res. 2002, 35, 746. (b) Jessop, P., Leitner,
W., Eds. Chemical Synthesis Using Supercritical Fluids; Wiley-VCH:
Weinheim, Germany, New York, 1999. (c) McHugh, M.; Krukonis, V. J.
Supercritical Fluid Extraction; Butterworth-Heinemann: Boston, 1994. (d)
Wells, S.; DeSimone, J. M. Angew. Chem., Int. Ed. 2001, 40, 519.
10.1021/jo060753p CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/27/2006
6432
J. Org. Chem. 2006, 71, 6432-6436

Baeyer-Villiger Oxidation in Supercritical CO2
solvating power, low surface tension, low viscosity, and high
diffusivity)3c confer interesting opportunities on this medium
to perform reactions with supported reagents and catalysts. For
instance, efficient nucleophilic substitution reactions have been
described with supported phase-transfer reagents,4 dehydration
of alcohols,5 and Friedel-Crafts alkylation6 over solid acid
catalysts, as has the oxidation of alcohols with silica-supported
chromium trioxide.7
FIGURE 1. Schematic view of the apparatus used in the oxidation
reactions: (A) CO2 cylinder; (B) diaphragm pump; (C) substrate
reservoir; (D) column charged with h-SiO2‚KHSO5; (E) micrometric
valve; (F) trap.
We herein report that hydrated potassium peroxomonosulfate
deposited onto silica, h-SiO2‚KHSO5, is an efficient reagent to
perform the Baeyer-Villiger oxidation of ketones into esters
or lactones in scCO2 under continuous-flow conditions. Hydra-
tion not only modifies the surface reactivity of the supported
reagent, but also suppresses acid-catalyzed side reactions which
are important competing pathways when the anhydrous reagent
is being used. The reactions were performed by flowing a
solution of the ketone in scCO2 through a column containing
the supported reagent and by recovering the product by
depressurization. Handling the peroxide under the reaction
conditions herein described simply involves placing a suitable
polypropylene container charged with the supported reagent into
the column and removing it once the reaction is complete. The
method significantly improves conventional Baeyer-Villiger
reaction procedures since it avoids the use of organic solvents
and reduces the physical operations required to isolate the
products to a simple depressurization. The reaction in scCO2 is
more efficient than in solution, and the results evidence the
solvating properties of scCO2.
FIGURE 2. Effect of scCO2 pressure (9) and flow (0) on the
conversion of ketone 1c at 40 °C.
Results and Discussion
Oxidation of ketones with h-SiO2‚KHSO5 in scCO2 was
carried out by flowing scCO2 (0.12 mL of scCO2 min-1) at 250
bar and 40 °C for 3.5 h through a reservoir containing the
substrate (0.8 mmol) and then through a column packed with
the supported reagent (molar ratio KHSO5:ketone ) 2:1) (Figure
1). The system was depressurized through a micrometric valve,
and the reaction products were collected in a trap cooled with
liquid nitrogen (Figure 1). Once the reaction was complete, the
consumed reagent recovered from the column had a white color
and loose appearance. GC analysis of the solution obtained by
a thorough washing of the recovered reagent with dichlo-
romethane allowed us to verify that organic material was not
retained on the solid. The results are shown in Table 1.
The results showed that cyclic ketones 1a-g efficiently react
with hydrated h-SiO2‚KHSO5 in scCO2 to exclusively provide
the corresponding lactones 2a-g with no hydrolysis of the
reaction products. Once the reaction was complete, iodometric
titration of the recovered supported reagent showed a nearly
stoichiometric conversion of the peroxide in relation to the
reacted ketone in most cases. The yields of the reactions reveal
the relative reactivities of the different ketones 1, which depend
on the ring size and the migratory ability of the substituents.
Increasing amounts of KHSO5 led to improved yields for the
less reactive ketones 1g and 1i. Thus, oxidation of acetophenone
(1i) with molar ratios KHSO5:ketone ) 2:1 and 4:1 enable
phenyl acetate (2i) to be obtained with 33% and 86% yields,
respectively, while decalone (1g) required 3 equiv of peroxide
to achieve a quantitative conversion of the substrate (entries 9
and 7, Table 1). For cycloheptanone (1d) and 2-octanone (1h),
however, an increase of the molar ratio KHSO5:1 did not
significantly improve the reaction yield (entries 4 and 8, Table
1).
Potassium peroxomonosulfate supported on silica was pre-
pared2 by mixing a ca. 2 M aqueous solution of the peroxide
with chromatographic-grade silica gel (particle size 0.040-0.060
mm) and by evaporating the solvent at room temperature under
vacuum. The resulting free-flowing white solid was either
lyophilized or dried under vacuum until a constant weight was
achieved. A peroxidic content of 0.5-2.2 mmol of peracid g-1
was determined in SiO2‚KHSO5 by iodometric titration. Deposi-
tion was quantitative in all cases. Acid-base titration of the
reagent gave a total acid content (KHSO4 + KHSO5) of 0.6-
2.3 mmol g-1, indicating a potassium hydrogen sulfate content
of 0.1-0.05 mmol g-1. The reagent can be stored at 2 °C in a
desiccator for weeks with no significant loss of peroxidic
content.
Prior to the reactions in scCO2, the supported reagent was
hydrated with 1.5 equiv of water in relation to the amount of
total acid (KHSO4 + KHSO5) on the silica. Hydration was
performed by placing two open vials containing the supported
reagent and the required amount of water respectively in a closed
chamber and by allowing the system to stand overnight at 4
°C. The reagent was used without further treatment. h-SiO2‚
KHSO5 can be stored at 4 °C for 2-3 weeks without significant
loss of peroxidic titer.
(4) DeSimone, J.; Selva, M.; Tundo, P. J. Org. Chem. 2001, 66, 4047-
4049.
(5) (a) Gray, W. K.; Smail, F. R.; Hitzler, M. G.; Ross, S. K.; Poliakoff,
M. J. Am. Chem. Soc. 1999, 121, 10711-10718. (b) Licence, P.; Gray, W.
K.; Sokolova, M.; Poliakoff, M. J. Am. Chem. Soc. 2005, 127, 293-298.
(6) Hitzler, M. G.; Smail, F. R.; Ross, S. K.; Poliakoff, M. Chem.
Commun. 1998, 359-360.
(7) Gonza´lez-Nu´n˜ez, M. E.; Mello, R.; Olmos, A.; Acerete, R.; Asensio,
G. A. J. Org. Chem. 2006, 71, 1039-1042.
J. Org. Chem, Vol. 71, No. 17, 2006 6433

Gonza´lez-Nu´n˜ez et al.
a
bar at a constant flow (0.12 mL of scCO2 min-1), temperature
(40 °C), and load of peracid (0.7 mmol of KHSO5 g-1), the
substrate conversion increased from 65% to 91%. However, a
further increase of up to 300 bar did not improve the reaction
efficiency (92%). Alternatively, when the flow of scCO2 was
increased from 0.13 to 0.33 mL of scCO2 min-1 at a constant
pressure (250 bar), temperature (40 °C), and load of peracid
(0.7 mmol of KHSO5 g-1), substrate conversion fell from 91%
to 57%. These results are apparently related to the substrate
residence time over the supported reagent, which diminishes
by either lowering the pressure at constant flow or increasing
the flow at a constant pressure.
TABLE 1. Oxidation of Ketones 1 with h-SiO2‚KHSO5 in scCO2
The load of peracid on h-SiO2‚KHSO5 also has a significant
effect on the reaction efficiency, although it depends on the
ketone used. Thus, at 250 bar and 40 °C with a flow of 0.12
mL of scCO2 min-1, the maximum conversion of 1b and
cyclohexanone (1c) was achieved with a load of 0.6-0.8 mmol
of peroxide g-1, yet the optimum load was 0.2-0.3 mmol of
KHSO5 g-1 in the case of the most reactive cyclobutanone (1a).
Evidently, the lower the load of the reagent, the lower the formal
concentration of the peroxide becomes at the surface of the solid
for a given molar ratio KHSO5:ketone. Alternatively, the higher
the amount of solid reagent required, the longer the path and
residence time of the substrate dissolved in scCO2 over the
reagent. An optimized load of peroxide in h-SiO2‚KHSO5 for a
given substrate would then be a compromise between the
opposite effects of the hydroperoxide concentration at the surface
and the residence time on the reaction rate. Also, the fact that
increasing the load of the reagent could lead the hydroperoxide
to occupy less accessible regions on the silica surface should
not be disregarded since this reduces its availability to react
with the flowing substrate.
a Reactions in scCO2 at 250 bar, 40 °C, and flow rate 0.11-0.15 mL of
scCO2 min-1, h-SiO2‚KHSO5 hydrated with 1.5 equiv of water with a load
of peroxide of 0.60-0.70 mmol of KHSO5 g-1, molar ratio ketone:KHSO5
) 1:2. b Esters or lactones were the only products (yield >99%). c Deter-
mined by GC analysis of the reaction mixtures, mass balance >98% in all
Hydration of SiO2‚KHSO5 is a critical parameter for a
successful reaction. We found that the oxidation of 1c with
anhydrous SiO2‚KHSO5 in scCO2 at 250 bar, 0.20 mL of scCO2
min-1, and 40 °C failed to produce any significant amount of
ꢀ-caprolactone (2c) in the trap. A thorough washing of the
supported reagent recovered from the column with dichlo-
romethane allowed isolation of (mass balance 75%) ꢀ-capro-
lactone (2c) (22%) and 2-cyclohexylidenecyclohexanone (78%).
A similar result was found2 for the reaction of cyclohexanone
(1c) with anhydrous SiO2‚KHSO5 in n-hexane. On the other
hand, hydration of SiO2‚KHSO5 with more than 2 equiv of water
in relation to the amount of total acid (KHSO4 + KHSO5) on
the silica led to a partial hydrolysis of the reaction products.
Hydroxyacids are not eluted by the flow of scCO2 since their
increased H-bonding ability with regard to esters provokes their
strong adsorption onto the acidic reagent surface. Thus, the
reaction of 1c with SiO2‚KHSO5 hydrated with 3 equiv of water
(molar ratio KHSO5:ketone ) 2:1) in scCO2 at 250 bar, 0.20
mL of scCO2 min-1, and 40 °C enabled us to recover only 46%
of the organic material and obtain only 88% of 2c and 5% of
unreacted ketone in the trap. Thoroughly washing the solid
reagent with dichloromethane allowed a recovery of 6-hydroxy-
hexanoic acid (7%).
the cases. d Flow rate 0.41 mL of scCO2 min-1 e Molar ratio ketone:KHSO5
.
) 1:3. f Molar ratio ketone:KHSO5 ) 1:5.
Recirculation of the scCO2 flow containing the mixture of
partially converted ketone and ester product over the supported
reagent allows for improving the conversion of the substrate.
A control experiment carried out by flowing a mixture of 1h
and the corresponding ester 2h (molar ratio 1h:2h ) 4:1) over
h-SiO2‚KHSO5 hydrated with 1.5 equiv of water and 0.65 mmol
of KHSO5 g-1 under our standard reaction conditions allowed
a further 20% conversion of the ketone 1h without significant
loss of the ester product 2h.
Reactions under flow-through conditions are generally less
effective than reactions in batch since the contact between the
substrate and immobilized reagent is limited to the residence
time of the flowing substrate solution over the solid reagent.
Accordingly, the use of a molar ratio KHSO5:cyclopentanone
(1b) ) 1:1 led to an incomplete conversion of the substrate.
On the other hand, the use of excess ketone 1b (molar ratio
KHSO5:1b ) 1:5) allowed a nearly complete conversion of the
supported peroxide and 20% conversion of ketone 1b into the
corresponding lactone 2b. In fact, this result indicates that all
the hydroperoxide groups at the active surface are available for
reacting with the substrate.
We have previously reported2 that surface-adsorbed potassium
hydrogen sulfate helps to catalyze the oxidation of ketones with
anhydrous SiO2‚KHSO5 in dichloromethane. The reaction
pathway can then be envisaged (Scheme 2) as an initial
adsorption equilibrium (I, II) of the substrate onto the acidic
surface. This is followed by a nucleophilic attack of a proximate
hydroperoxide group to the carbonyl group of an acid-base
Pressure and flow effects on substrate conversion are shown
in Figure 2. When the pressure was increased from 100 to 250
6434 J. Org. Chem., Vol. 71, No. 17, 2006

Baeyer-Villiger Oxidation in Supercritical CO2
SCHEME 2. Reaction Mechanism at the Active Surface
complex (II, III) to provide the tetrahedral intermediate, which
then undergoes transposition of the alkyl chain (IV, V).
The relative position of the adsorption equilibrium is deter-
mined by the polarity and solvating ability of the solvent. Apolar
solvents favor strong adsorption of the ketone onto the acidic
centers of the active surface8 and reduce the possibility of the
substrate reaching a reactive hydroperoxide group. The acid-
base complex ketone-KHSO4 becomes a new reactive species
immobilized onto the active surface which then has the
opportunity to condense with further ketone molecules carried
by the flowing scCO2 solution. It is worth noting that the
oxidation reaction produces KHSO4 as the reduced species of
the peracid.
The dramatic effect of hydration of SiO2‚KHSO5 on the
efficiency and selectivity of the reaction in scCO2 can then be
attributed to a lowering of the acidic strength of surface-bonded
KHSO4 by coordination with water. In this way, substrate
adsorption at the acidic sites of the solid surface becomes weaker
and the ketone molecules have more opportunities to find
reactive hydroperoxide groups at different regions of the solid
surface. In addition, the fact that hydration forms a polar and
protic layer over the reagent surface should not be disregarded,
which extends the availability of acidic protons farther from
the anchoring center of potassium hydrogen sulfate and increases
the efficiency of acid catalysis at the reagent surface.
Comparison of scCO2 and n-Hexane under Flow-Through
Reaction Conditions. To compare the solvating properties of
scCO2 and n-hexane in the Baeyer-Villiger reaction with
hydrated h-SiO2‚KHSO5, we performed the reaction of 1c in
n-hexane under flow-through conditions. We previously estab-
lished the average concentration of 1c in scCO2 under our
standard reaction conditions by placing a weighed amount of
the ketone in the substrate reservoir, flowing scCO2 at 250 bar,
40 °C, and 0.12 mL min-1 for 90 min, and determining the
amount of ketone which remained in the reservoir after
depressurization. The experiment was repeated four times with
different amounts of 1c, and a 0.027 ( 0.002 M average
concentration of 1c in scCO2 was yielded under our standard
reaction conditions.
The reaction of 1c with hydrated h-SiO2‚KHSO5 in n-hexane
under flow-through conditions was carried out by flowing a 0.03
M solution of the ketone in n-hexane at 0.106 mL min-1 through
a column containing the supported reagent (molar ratio ketone:
KHSO5 ) 1:2), which was heated at 40 °C. The solution was
collected in an ice-cooled flask placed at the end of the column.
We could recover only 70% of the organic material, 70% of
which was recovered in the collector and 30% from the washings
of the solid support with dichloromethane. GC analysis of the
mixture showed unreacted ketone (45%), ꢀ-caprolactone (33%),
and 2-cyclohexylidenecyclohexanone (22%).
These results show that the reaction in scCO2 is much more
efficient than in n-hexane. This can be attributed to the transport
properties of scCO2,3 whose high diffusivity, low viscosity, and
low surface tension allow for the substrate solution to penetrate
the porous matrix better and to reach a more extensive active
surface of the solid reagent than n-hexane. However, the results
also evidence a better solvating ability of scCO2 compared to
n-hexane since the former completely removes the reaction
products from the acidic solid support and avoids the formation
of aldol condensation side products. The association of either
the ketone or ester with the acidic sites at the supported-reagent
surface is less favorable in scCO2 than in n-hexane, which
should be attributed to a better solvation of polar oxygenated
molecules in the supercritical medium. These results indicate
that scCO2 has peculiar solvating properties which cannot be
compared so easily to those of hydrocarbons.
In short, we have shown that hydrated h-SiO2‚KHSO5 is an
efficient reagent to perform the Baeyer-Villiger reaction of a
variety of ketones in scCO2 under flow-through conditions. The
reactions were performed by flowing a solution of the ketone
in scCO2 through a column containing the supported reagent
and recovering the product by depressurization. Handling the
peroxide under the reaction conditions herein described simply
involves placing a suitable polypropylene container charged with
the supported reagent into the column and removing it once
the reaction is complete. This method, which requires high-
pressure equipment, avoids the use of organic solvents as well
as the contamination of products with the reduced form of the
oxidant and simplifies the physical operations required to isolate
reaction products. The efficiency of the method depends on the
reactivity of the substrate and its solubility in scCO2, which
determine the consumption of CO2 and duration of the experi-
ment. We have shown that the surface reactivity of the supported
reagent can be modulated by hydration, which diminishes the
strength of the acidic sites of the active surface. The results
also show that the transport and solvating properties of scCO2
are significantly better than those of n-hexane in these reactions.
Experimental Section
General Procedures. Aqueous solutions of peroxomonosulfate
and SiO2‚KHSO5 were prepared following reported procedures.2
Peroxidic and total acid contents of the supported reagent were
determined by iodometric and acid-base titrations. The amount
of KHSO4 in the supported reagent was determined by the
difference found between the total acid and peroxidic contents.
Hydration of SiO2‚KHSO5 was performed by placing in a 50 mL
(8) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry,
2nd ed.; VCH: Weinheim, Germany, 1988.
J. Org. Chem, Vol. 71, No. 17, 2006 6435

Gonza´lez-Nu´n˜ez et al.
closed chamber two vials containing, respectively, a weighed sample
of SiO2‚KHSO5 and 1.5 equiv of H2O with regard to the total acid
content in the supported reagent, and the system was left to stand
overnight at 4 °C.
and 13C NMR. The solid reagent recovered from the column was
washed four times with 15 mL of dichloromethane in a round-
bottomed flask under magnetic stirring. The filtered solution was
analyzed by GC and then evaporated under vacumm at 0 °C.
The high-pressure equipment consisted of a 250 mL AISI 316
stainless steel jacketed reactor, a high-pressure micrometric valve
placed at the outlet of the reactor, a diaphragm pump (Orlita MHS
30/8) with a maximum theoretical flow of 8.44 L/h of liquid CO2,
and a set of HIP high-pressure valves, pressure and temperature
probes, and security rupture disks suitably placed to control the
flow of CO2 along the system.
Oxidation of Ketones 1 to Esters 2 with h-SiO2‚KHSO5 in
scCO2. General Procedure. Cyclohexanone (1c) (0.85 mmol) was
placed in a Teflon cylinder covered at the top with a screw cap
with two 1/4-28 UNF threaded ports. The reservoir was kept open
in one of the ports, while the second was connected to a 5 mL
Rezorian Luer-lock syringe-tip cartridge (Supelco) charged with
2.5 g of h-SiO2‚KHSO5 (1.7 mmol) by 1/8 in. Teflon tubing and
suitable fittings. Two filters placed at the bottom and the top of
the cartridge prevented any displacement of the solid reagent. The
assembly was placed within the 250 mL autoclave thermostated at
40 °C, and the column outlet was connected to the reactor outlet
by Teflon tubing and a septum. The high-pressure micrometric valve
outlet was connected to a trap cooled with a liquid nitrogen bath
by 1/8 in. Teflon tubing. The pressure in the trap was equilibrated
with a flow of nitrogen.
Oxidation of 1c with h-SiO2‚KHSO5 in n-Hexane under Flow-
Through Conditions. A 5 mL Rezorian Luer-lock syringe-tip
cartridge (Supelco) was charged with 2.5 g of h-SiO2‚KHSO5 (1.75
mmol) and placed within a vertical jacketed cylinder heated at 40
°C by a circulating bath. The cartridge outlet was connected to a
100 mL flask cooled at 0 °C by 1/8 in. Teflon tubing. A 50 mL
polypropylene syringe was charged with 30 mL of a 0.03 M solution
of 1c in n-hexane, placed in a syringe pump set at 0.106 mL min-1
,
and connected to the cartridge inlet by 1/8 in. Teflon tubing and
suitable fittings. The assembly was operated until the syringe was
emptied. Then the cartridge was washed with an additional 30 mL
of n-hexane. The solution in the collector was analyzed by GC,
and the solvent was removed under vacuum at 0 °C. The solid
reagent recovered from the column was washed four times with
15 mL of dichloromethane in a round-bottomed flask under
magnetic stirring. The filtered solution was analyzed by GC and
then evaporated under vacuum at 0 °C.
Acknowledgment. Financial support via grants from the
Spanish Direccio´n General de Investigacio´n (BQU2003-3005)
and Generalitat Valenciana (Grupos 2001-2003 and GV04B-
161) is gratefully acknowledged. A.O. thanks the Spanish
Ministerio de Educacio´n y Ciencia for a fellowship. We thank
SCSIE (UVEG) for access to their instrumental facilities.
Dedicated to Prof. Marcial Moreno-Man˜as, in memoriam.
The reactor was closed, charged with CO2, and then pressurized
to 250 bar. This operation usually required ca. 15 min. The stroke
volume of the pump and the aperture of the high-pressure
micrometric valve at the reactor outlet were regulated to achieve
steady continuous-flow conditions at 250 bar. The CO2 flow at the
system outlet was monitored with a bubble flow meter.
Supporting Information Available: Gas chromatograms and
NMR spectra of the reaction crudes. This material is available free
of charge via the Internet at http://pubs.acs.org.
The system was left to operate for 3.5 h and then completely
depressurized through the micrometric valve. The trap was allowed
to warm to room temperature. The colorless residue was dissolved
1
in deuterated chloroform and analyzed by GC, GC-MS, and H
JO060753P
6436 J. Org. Chem., Vol. 71, No. 17, 2006