Impact of carbon dioxide pressurization on liquid phase organic reactions: A case study on Heck and Diels-Alder reactions

Article abstract of DOI:10.1002/adsc.200800212

Heck coupling reactions of methyl acrylate with various aryl bromides have been investigated using a Pd/TPP catalyst in toluene under pressurized CO ₂ conditions up to 13 MPa. Although CO₂ is not a reactant, the pressurization of the reaction liquid phase with CO₂ has positive and negative impacts on the rate of Heck coupling depending on the structures of the substrates examined. In the case of either 2-bromoacetophenone or 2-bromocinnamate, the conversion has a maximum at a CO₂ pressure of about 3 MPa; for the former, it is much larger by a factor of 3 compared with that under ambient pressure. For 2-bromobenzene, in contrast, the conversion is minimized at a similar CO₂ pressure, being half compared with that at ambient pressure. In the other substrates, including the other isomers of these three aryl bromides, the conversion simply decreases or does not change so much with the CO₂ pressure. To examine the factors responsible for the effects of CO₂ pressurization, the phase behavior and the molecular interactions with dense phase CO₂ have also been studied by visual observation and in situ high pressure FT-IR spectroscopy. In addition, impact of CO₂ pressurization was also studied for the Diels-Alder reactions of isoprene with a few dienophiles like methyl acrylate, methyl vinyl ketone, and acrolein in the same solvent, toluene, but a heterogeneous silica-alumina catalyst was used (the reaction system was liquid-solid biphasic). When the CO₂ pressure is raised, the conversion monotonously decreases for the three dienophiles; however, the product selectivity changes with the pressure, in particular for acrolein. The FT-IR spectroscopic measurements suggest that its reactivity is altered by interactions with CO₂ molecules under pressurized conditions.

Full text of DOI:10.1002/adsc.200800212

FULL PAPERS
DOI: 10.1002/adsc.200800212
Impact of Carbon Dioxide Pressurization on Liquid Phase
Organic Reactions: A Case Study on Heck and Diels–Alder
Reactions
S. Fujita,a T. Tanaka,a Y. Akiyama,a K. Asai,a J. Hao,b F. Zhao,b and M. Araia,*
a
Division of Chemical Process Engineering, Graduate School of Engineering, Hokkaido University, Sapporo 060-8628,
Japan
Fax : (+81)-(0)11-706-6594; e-mail: marai@eng.hokudai.ac.jp
State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of
b
Sciences, Changchun 130022, Peopleꢀs Republic of China
Received: April 6, 2008; Revised: May 8, 2008; Published online: June 23, 2008
Abstract: Heck coupling reactions of methyl acrylate lecular interactions with dense phase CO2 have also
with various aryl bromides have been investigated been studied by visual observation and in situ high
using a Pd/TPP catalyst in toluene under pressurized pressure FT-IR spectroscopy. In addition, impact of
CO2 conditions up to 13 MPa. Although CO2 is not a CO2 pressurization was also studied for the Diels–
reactant, the pressurization of the reaction liquid Alder reactions of isoprene with a few dienophiles
phase with CO2 has positive and negative impacts on like methyl acrylate, methyl vinyl ketone, and acro-
the rate of Heck coupling depending on the struc-
A
tures of the substrates examined. In the case of ous silica-alumina catalyst was used (the reaction
either 2-bromoacetophenone or 2-bromocinnamate, system was liquid-solid biphasic). When the CO2
the conversion has a maximum at a CO2 pressure of pressure is raised, the conversion monotonously de-
about 3 MPa; for the former, it is much larger by a creases for the three dienophiles; however, the prod-
factor of 3 compared with that under ambient pres- uct selectivity changes with the pressure, in particu-
sure. For 2-bromobenzene, in contrast, the conver- lar for acrolein. The FT-IR spectroscopic measure-
G
sion is minimized at a similar CO2 pressure, being ments suggest that its reactivity is altered by interac-
half compared with that at ambient pressure. In the tions with CO2 molecules under pressurized condi-
other substrates, including the other isomers of these tions.
three aryl bromides, the conversion simply decreases
or does not change so much with the CO2 pressure. Keywords: carbon dioxide; Diels–Alder reaction; ex-
To examine the factors responsible for the effects of panded solvent phase; Heck reaction; multiphase re-
CO2 pressurization, the phase behavior and the mo- action; pressure effect
Introduction
The present authors also studied the selective hy-
drogenation of a,b-unsaturated aldehydes of cinna-
A C O-expanded organic liquid phase is an interesting maldehyde and citral in CO2-exapnded liquid phases
2
medium for catalytic reactions; it is produced by in- using soluble metal complexes and insoluble support-
creasing the volume of an organic liquid through the ed metal particles as catalysts.[4] For the former, for
dissolution of a relatively large amount of CO2 under instance, the total conversion is enhanced by the pres-
high pressure conditions. Subramaniam and his co- surization of the liquid substrate phase with CO2 at
workers first demonstrated the merits of using CO2- pressures up to 16 MPa. This is the same positive
expanded organic solvents for homogeneous catalytic effect as observed so far in other chemical reactions
oxidation reactions.[1] The CO2-expanded liquid including gaseous reactants such as H2, O2, and CO,
phases offer several advantages like increased solubil- which may be similarly ascribed to the increase of the
ity of gaseous reactants involved, enhanced reaction H2 concentration in the liquid substrate phase. Fur-
rates, and decreased volume of organic solvents used. thermore, CO2 pressurization shows another interest-
Studies have also reported on other catalytic reactions ing feature, namely, the selective hydrogenation of
of hydrogenation[2] and hydroformylation.[3]
the C=O bond can be promoted compared with that
of the C=Cbond, resulting in an increase in the selec-
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1615

S. Fujita et al.
FULL PAPERS
tivity towards unsaturated alcohol (cinnamyl alcohol dense phase CO2 (solvent) using an SiO2 catalyst and
in the case of cinnamaldehyde). An important factor showed the yield to simply decrease with the CO2
responsible for this promotion should be an interac- pressure.[7a] In our liquid-solid reactions of isoprene
tion of the dissolved CO2 molecules with the sub- with a few carbonyl compounds under pressurized
strate, in particular with its carbonyl group, through CO2, however, the selectivity towards the [2] and [3]
which the reactivity of the carbonyl group may be adducts can be altered by the CO2 pressure. The re-
modified. These interactions were evidenced by in sults of Diels–Alder reactions at elevated static pres-
situ high pressure FT-IR measurements.[4b] We may sures are reviewed in a comprehensive book.[7e] Ele-
expect from these results that the CO2 pressurization vated static pressures (100–2000 MPa used in the pre-
has not only a physical effect of enhancing the disso- vious works) are expected to enhance the rate of a re-
lution of gaseous reactants but also a chemical effect action that exhibits
a large negative activation
of improving the reactivity of substrates and/or cata- volume, such as the Diels–Alder reaction and the
lysts, in addition to modification of solvent properties Heck coupling.[6,7e] However, the present reaction sys-
(polarity, for example).
tems are organic liquid phases pressurized by gaseous
Thus, it is interesting to examine the effects of CO2 CO2 at not such elevated pressures, i.e., below
pressurization on true liquid phase reactions that do 20 MPa.
not include any gaseous reactants. In the present
work, our attention has been directed to Heck cou-
pling using molecular homogeneous catalysts, which is Results and Discussion
one of most important carbon-carbon bond forming
reactions in organic synthesis.[5] The influence of such Heck Reactions
an elevated pressure of 100 MPa or above was previ-
ously studied on the rate of reaction and the product Previous Heck Reactions under Conventional
selectivity in several organic reactions including cross- Conditions: Influence of Dilution
coupling reactions. The interesting results and discus-
sions obtained so far have been described in previous As described later, the reactions were observed to
reviews.[6] In those works chemical reactions were in- occur in an expanded solvent (toluene) phase in the
vestigated under greatly elevated static pressures. In presence of pressurized CO2. The volume expansion
the present work, the liquid reaction mixture is pres- of the reaction phase causes the dilution of reacting
surized by CO2 at lower pressures, below 20 MPa, and species and this will change the rate of reaction. Since
a simple static pressure effect cannot be expected. the analysis of Heck reaction kinetics is complicated,
However, the CO2 pressurization is expected to have only a few workers[8] have attempted so far to formu-
physical and/or chemical effects. The volume of a late the rate of reaction in spite of its importance in
liquid phase may expand by dissolution of CO2 mole- practical organic synthesis. Previously the authors pre-
cules, causing a dilution of reacting species in the sented an empirical rate equation for the Heck cou-
liquid phase. The dissolved CO2 molecules may have pling of methyl acrylate and iodobenzezne in NMP
interactions with substrates and/or catalysts, modify- with the same catalyst and organic base as used in the
ing their reactivity and activity. This work is the first present work.[8a] This rate equation can be used to es-
case study to investigate the effects of CO2 pressuriza- timate the influence of simple dilution on the rate of
tion on liquid phase reactions. The Heck reactions of Heck reaction under atmospheric pressure. This esti-
methyl acrylate with various aryl bromides were mation will be helpful for later discussion of the Heck
tested with a homogeneous Pd/TPP catalyst in tolu- reactions under CO2 pressurization, although the sub-
ene under pressurized CO2 conditions. The effects of strates and reaction conditions are different from
CO2 pressurization will be compared and discussed those used in the present work. The rate of reaction
with respect to the CO2 pressure and the molecular was first calculated under such concentration condi-
structure of the substrates. This work shows that the tions: methyl acrylate, iodobenzene, triethylamine
rate of liquid phase Heck reaction is indeed enhanced 1 mmolcmÀ3 each; Pd(OAc)2 0.01 mmolcmÀ3; TPP
A
by the presence of pressurized CO2, although the rate 0.04 mmolcmÀ3. The rate of the reaction is further
enhancement is limited to bromoarenes of a certain calculated for smaller concentration conditions, sup-
structure. In addition, the impact of CO2 pressuriza- posing that the reaction mixture is diluted by adding
tion was also studied for the liquid-solid biphasic the solvent. Figure 1 gives the relative rate of the
Diels–Alder reactions of isoprene with a few carbonyl Heck reaction with respect to the one at the initial
compounds using a heterogeneous silica-alumina cata- concentrations as a function of the extent of dilution
lyst. Unfortunately, the CO2 pressurization was ob- (the initial concentration against the diluted concen-
served to merely decrease the rates of Diels–Alder re- tration). As expected, the rate of reaction tends to de-
actions. Tester et al. investigated the Diels–Alder re- crease simply with the dilution (caused by the expan-
actions of methyl vinyl ketone and penta-1, 3-diene in sion of the reaction volume in the following cases
1616
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1615 – 1625

Impact of Carbon Dioxide Pressurization on Liquid Phase Organic Reactions
FULL PAPERS
Figure 1. Change in the rate of Heck coupling of iodobenzene and methyl acrylate with the extent of dilution of the liquid
phase (NMP solvent) as estimated from an empirical rate expression obtained previously.[8a] (a) Relative rate per unit
volume with respect to that with no dilution (under initial concentration conditions) at 343 K (mole volumeÀ1 timeÀ1); (b)
Relative rate produced by extent of dilution=total rate in expanded volume (mole timeÀ1). The initial concentrations of io-
dobenzene, methyl acrylate, triethylamine, palladium, and TPP used for this calculation are given in the text.
with CO2 pressurization), in particular in the range of version tends to simply decrease with pressure; no
the small extent of dilution up to a factor of 2.
positive effect was observed when the liquid reaction
mixture was pressurized by CO2. The results with acti-
vated bromobenzenes are given in Figure 3, which are
complicated depending on the aryl bromide substrate
and its structure. It is interesting to note that the con-
Heck Reactions under Pressurized CO2 Conditions
The pressurization with CO2 was observed to have version increases with the pressure in the range up to
different impacts depending on the aryl bromides 3 MPa for 2-bromoacetophenone and ethyl 2-bromo-
used. Figure 2 shows the results of the conversion of cinnamate. For the former substrate, the maximum
iodobenzene and non-activated bromobenzene conversion is larger by a factor of about 3 compared
against CO2 pressure. For both the substrates the con- with that under ambient pressure. For the other iso-
mers of these two substrates, the pressurization with
CO2 causes a significant or a marginal decrease in the
conversion, similar to the results of the largest sub-
strates, ethyl bromocinnamates. The conversion of 4-
bromobenzaldehyde changes little with the CO2 pres-
sure, while that of 2-bromobenzaldehyde is minimized
at around 2 MPa. These results demonstrate that the
CO2 pressurization can accelerate the conversion of
the Heck reaction for a few specific substrates.
To examine the importance of CO2, N2 was used at
pressures up to 6 MPa instead of CO2 for Heck reac-
tions of a few substrates (2-bromoacetophenone, ethyl
2-bromocinnamate, and 2-bromobenzaldehyde) for
which the CO2 pressurization has significant positive
or negative impacts. The pressurization of these liquid
reaction mixtures with N2 was found to have no
impact on the reactions. That means, the chemical
nature of CO2, as well as high pressure, is an impor-
tant factor responsible for the changes of the Heck
conversion observed (Figure 3).
Figure 2. Influence of CO2 pressurization on the Heck reac-
tion of iodobenzene and bromobenzene with methyl acry-
late. Reaction conditions: bromobenzene (iodobenzene)
10 mmol, methyl acrytate 10 mmol, Pd(OAc)2 0.1 mmol,
A
TPP 0.4 mmol, triethylamine 10 mmol, toluene 10 cm3, tem-
perature 393 K, time 2 h.
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1617

S. Fujita et al.
FULL PAPERS
Figure 3. Influence of CO2 pressurization on the Heck reactions of 2-, 3-, 4-isomers of bromobenzaldehydes, bromoacetophe-
nones, methyl trans-bromocinnamate, and ethyl trans-bromocinnamate with methyl acrylate. Reaction conditions: bromo-
A
N
Volume Expansion of Organic Solvent Phase by CO2
Pressurization
HighPressure FT-IR of Substrates in Pressurized
CO2
At high pressures a large quantity of CO2 is dissolved According to the established reaction mechanisms,
into a liquid phase and this causes an expansion of the Heck reaction goes through several reaction
the liquid phase.[1] The volume expansion of the sol- steps.[5] We are at present unable to explain which
vent phase by CO2 pressurization was examined at steps are influenced by the CO2 molecules dissolved
the reaction temperature (393 K) by visual observa- in the reaction solution. We tested the Heck reactions
tion of pure toluene solvent at different CO2 pres- of various aryl bromides with methyl acrylate using
sures. Figure 4 (a) shows that the volume of the tolu- triethylamine in toluene. The interaction of CO2 with
ene phase increases with pressure by dissolution of triethylamine is likely to be weak.[9] High pressure
CO2 molecules. The extent of volume expansion with FT-IR measurements were used to examine the inter-
respect to the initial volume at atmospheric pressure actions of dense phase CO2 with a few selected aryl
is given in Figure 4 (b). The volume expansion was bromides, as well as methyl acrylate, for which signifi-
also examined for toluene in the presence of sub- cant effects of CO2 pressurization were observed. Un-
À
strates (methyl acrylate and 2-bromoacetophenone or fortunately it is difficult to examine the C Br bond of
2-bromobenzaldehyde) and base (triethyleamine) at aryl bromides, for which the corresponding absorption
similar concentrations to those used in the Heck reac- bands may appear at small wavenumbers of 600–
tions. The extent of volume expansion was not ob- 500 cmÀ1,[10] due to the limitation of the FT-IR appa-
served to change with the presence of these reacting ratus used. Figure 5 shows FT-IR spectra in the range
species. In addition, no volume expansion was ob- of the n
served when the liquid phase was pressurized by N2 nones (see Figure 3) in dense phase CO2. The n
instead of CO2 at similar pressures. absorption peak position for 2-bromacetophenone
ACHTREUNG
ACHTREUNG
1618
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1615 – 1625

Impact of Carbon Dioxide Pressurization on Liquid Phase Organic Reactions
FULL PAPERS
Figure 4. Volume expansion of pure toluene with pressurized CO2 at different pressures at 393 K using an initial volume of
toluene of 23.8 cm3 in a reactor volume 85 cm3.
Figure 5. FT-IR spectra of 2-, 3-, 4-bromoacetophenones (a, b, c) dissolved in CO2 at different pressures of 0.1, 3, 6, 8, 10,
and 12 MPa (from bottom to top) and at 393 K. For b, the absorption at 0.1 MPa was very weak and so it is increased by
three times herein.
changes with CO2 pressure in a different manner as bromoacetophenone. The above discussion is based
compared with those of 3- and 4-bromoacetophe- on the FT-IR results of the substrates dissolved in
nones (Figure 6). The n(C=O) absorption band of the dense CO2 phase. However, the same discussion may
G
former substrate in dense phase CO2 appears at larger be possible for the CO2-dissolved expanded toluene
wavenumbers by 7–10 cmÀ1 than those of the latter phase, where the Heck reactions occurred, because
two ones; the carbonyl bond of 2-bromoacetophenone the concentration of CO2 in this phase is large.
is stronger compared with that of 3- or 4-bromoaceto-
We also made similar FT-IR measurements with
À
phenone. It is thus suggested that the C Br bonds of methyl acrylate. Unfortunately, the solubility in dense
the three substrates are modified by the presence of phase CO2 was decreased with increasing CO2 pres-
dense CO2, indirectly through interactions of CO2 sure, and so noisy spectra were obtained. It was thus
À
with the other moieties or directly with the C Br difficult to examine the influence of dense CO2 mole-
moiety. In addition, the modification of 2-bromoace- cules on the strength of the bonds of this substrate.
tophenone should be different from that of 3- or 4- However, we may speculate that CO2 is unlikely to
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1619

S. Fujita et al.
FULL PAPERS
IR measurements indicate that dense phase CO2 mol-
ecules interact with the aryl bromide substrates in dif-
ferent manners depending on their structures. It is
speculated that the steps involved with a bromoarene
are influenced by the CO2 molecules dissolved in the
liquid phase. The Heck reaction is such a complicated
system that it goes through several steps with several
reacting species (Pd/TPP catalyst and substrates),[5]
and so further discussion is not possible at present.
We think that spectroscopic measurements of the re-
action mixtures under the reaction conditions are also
required.
Figure 6. Dependence of n(C=O) absorption peak position
N
~
*
on CO2 pressure for 2-, 3-, 4-bromoacetophenones ( , &,
)
Diels–Alder Reactions
based on the FT-IR results shown in Figure 5.
The impact of CO2 pressurization was also studied for
have a positive impact on the reactivity of methyl a different reaction system, the Diels–Alder reaction
acrylate because of its decreasing solubility in dense with a heterogeneous SiO2·Al2O3 catalyst, which was
phase CO2 (decreasing attractive interactions with a liquid-solid biphasic system including no gaseous re-
ACHTREUNG
CO2).
actant. We did not use Heck reactions using heteroge-
neous Pd catalysts like Pd/Cbecause the leaching of
supported Pd species into the solvent would occur
and the reactions would not be heterogeneous.[5,15]
The Diels–Alder reactions were also studied at such
Roles of Dense CO2 Molecules in Heck Reactions
It should be noted again that the rate of the liquid an elevated static pressure of 1000 MPa.[7e] Such ele-
phase Heck reaction can be enhanced by pressurizing vated pressures are expected to accelerate the rate of
the reaction mixture with CO2 at a low pressure of 2– reaction because the Diels–Alder reaction as well as
4 MPa, although the enhancement effect is limited to the Heck coupling may exhibit a negative activation
a few bromoarenes of a certain structure (2-bromo- volume. The present reactions were conducted under
A
discussion is given here on the role of dense phase
CO2 in this observed enhancement. The pressuriza-
tion with CO2 decreases the concentration of reacting Diels–Alder Reactions under Pressurized CO2
species in the liquid reaction phase as a result of its conditions
volume expansion due to the dissolution of CO2 at
high pressures (Figure 1 and Figure 4). This is a nega- Figure 7 shows the results obtained with three differ-
tive impact on the rate of reaction, and so the other ent dienophiles. Unfortunately, the conversion was
positive impacts involved with CO2 should be superior observed to merely decrease with the CO2 pressure.
to that negative impact.
This simple decrease of the conversion should result
Toluene is a non-polar solvent with a dielectric con- from a simple dilution effect of CO2 molecules into
stant er =2.38 and an empirical parameter of solvent the liquid phase at high pressures. One may consider
polarity ET(30)=33.9 kcalmolÀ1 at 298 K.[11] Dense another possibility that CO2 molecules are adsorbed
CO2 fluid is a non-dipolar solvent[12] and the er value on the surface of the SiO2·Al2O3 catalyst and this may
increases with pressure, ranging from 1.020 to 1.556 at mask the catalytically active sites and decrease the
298 K at pressures of 2–18 MPa.[13] The ET(30) value total activity. A few previous authors reported de-
was also estimated for dense CO2 fluid, being 28.5 tailed investigations of the adsorption of CO2 on a
and 29.0 kcalmolÀ1 at 15 and 20 MPa, respectively, at similar oxide, SiO2 and C.[7a,16]
313 K.[14] Then, it is assumed that the dissolution of
Table 1 summarizes the product selectivity at differ-
CO2 into the toluene phase scarcely changes its ent CO2 pressures in the Diels–Alder reactions of iso-
nature as a continuum of non-polar reaction medium. prene with the three dienophiles. The results show
Several aryl bromides and methyl acrylate were that the ratio of [3] adduct to [2] adduct can be al-
used in the present Heck reactions and the effects of tered by the CO2 pressure, which changes most exten-
CO2 pressurization significantly depended on the sively from 47:53 to 73:27 at pressures up to 16 MPa
structures of bromoarenes used. No effects were ob- for acrolein. The change is less significant for the
served with N2 pressurization and so the molecular other dienophiles but the ratio does also change from
features of CO2 are important. The high pressure FT- 63:37 to 73:27 and from 70:30 to 75:25 for methyl
1620
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1615 – 1625

Impact of Carbon Dioxide Pressurization on Liquid Phase Organic Reactions
FULL PAPERS
Table 1. Influence of CO2 pressure on the selectivity of
liquid-solid biphasic Diels–Alder reactions of isoprene with
methyl acrylate, methyl vinyl ketone, and acrolein.[a]
CO2 pres
A
Product selectivity (3:2 ratio)
[MPa]
Methyl
acrylate
Methyl vinyl
ketone
Acrolein
0.1
4
70:30
63:37
66:34
69:31
70:30
73:27
47:53
(71:29)[b]
70:30
(68:32)[b]
55:45
(71:29)[b]
70:30
(69:31)[b]
69:31
8
(71:29)[b]
76:24
(83:17)[b]
70:30
12
16
(74:26)[b]
75:25
(83:17)[b]
73:27
(76:24)[b]
(83:17)[b]
[a]
Reaction conditions: SiO2·Al2O3 catalyst 0.2 g, isoprene
20 mmol, CH2=CH-R 10 mmol, toluene 4 cm3, tempera-
ture 353 K, time 2 h (4 h for methyl acrylate).
[b]
Data collected with ethanol instead of toluene.
Figure 7. Influence of CO2 pressurization on the Diels-Alder
reactions of isoprene with CH2 =CH-R in toluene (a) and
ethanol (b). Reaction conditions: SiO2·Al2O3 0.2 g, isoprene
20 mmol, CH2=CH-R 10 mmol, solvent 4 cm3, temperature
353 K, time 2 h (4 h for methyl acrylate).
uene would affect the reactivity of substrates; they
surround the substrate molecules inducing interac-
tions between the substrate and CO2 molecules in the
liquid phase. As the pressure is raised, the state of
vinyl ketone and methyl acrylate, respectively, in the such a clump of substrate-CO2 molecules in an organ-
same pressure range. In other words, the ratio of 3 to ic solvent of toluene may become similar to that in a
2 and the effect of CO2 pressure on this ratio are dense phase CO2 medium. Then, the product selectiv-
larger in the order of acrolein>methyl vinyl ketone> ity obtained under highly pressurized CO2 should be
methyl acrylate. At a high pressure of 16 MPa, very similar to that in the CO2 medium. It is known that
similar 3 to 2 ratios are seen for the three dienophiles, the selectivity of Diels–Alder reactions is different in
which are similar to those with Diels–Alder reactions a bulk organic phase and in organic droplets (dis-
of cyclopentadiene with methy vinyl ketone and persed in water), in which water-insoluble reactants
methyl acrylate in dense phase CO2 reported by are clumped together, increasing the endo selectivity
Tester et al.[7a] They reported that the product selec- in the reaction of cyclopentadiene and methyl vinyl
tivity was not affected by CO2 pressure. It should be ketone.[17]
noted again that the selectivity of liquid phase Diels–
The reaction runs were also conducted with methyl
Alder reactions may be changed with CO2 pressure acrylate or acrolein in ethanol instead of toluene
when these are conducted under pressurized CO2 (in (Figure 7, Table 1). Although toluene and ethanol are
CO2-dissolved expanded solvent phase). The Diels– different in solvent properties,[11] the results of Diels–
Alder reactions in dense phase CO2 were previously Alder reactions in these organic solvents are similar
investigated by several workers and the effects of for the ester methyl acrylate. For the aldehyde acrole-
CO2 pressure on the total conversion and the product in, however, the conversion in ethanol and the 3 to 2
selectivity were observed for some reactions, as de- ratio are larger as compared with those in toluene.
scribed in review articles.[7b–d] As discussed above for The enhancement of Diels–Alder reactions by solvent
Heck reactions, the pressurization and dissolution of effects in hydrogen-bonded donor (HBD) solvents is
CO2 in the toluene phase will scarcely change its fea- known with carbonyl-containing reactants and the hy-
tures as a continuum of solvent for the Diels–Alder drogen bonding between the substrate and the sol-
À
À À
reactions as well. The CO2 molecules dissolved in tol- vent, C=O···H O R, is significant in accelerating
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1621

S. Fujita et al.
FULL PAPERS
the reactions.[11] Such a hydrogen bond may also for qualitative discussion of interactions with CO2
cause the increase of the 3 to 2 ratio in ethanol. The 3 molecules. The strength of the absorption bands was
to 2 ratio is increased with increasing CO2 pressure in not observed to increase monotoneously with CO2
either ethanol or toluene and the extent of change is pressure, and an important factor for this should be a
large at around the critical pressure of pure CO2. The change of absorption coefficient with the pressure.[18b]
CO2 molecules dissolved in the solvents should play Our attention will herein be given to the absorption
important roles for determining the product selectivi- peak positions and so the spectra of Figure 8 are
ty as well as the total conversion. The product selec- given at appropriate different absorption scales. For
tivity may be altered by interactions of the carbonyl acrolein, the spectra at 0.1 MPa and 4 MPa are differ-
group with the HBD solvent (ethanol in the present ent but the absorption occurs in the same wavenem-
case) or the dense phase CO2 molecules (vide infra). ber range. A weak band of n
A
1650 cmÀ1 and the n
AHCTREUNG
bands, of which a stronger band is located at
1715 cmÀ1 and a weaker band at around 1740 cmÀ1.
The absorption band splitting was previously reported
for a few carbonyl compounds.[18] The weaker band
HighPressure FT-IR of Substrates in Pressurized
CO2
In situ high pressure FT-IR was used to examine mo- seems to make a blue-shift with increasing CO2 pres-
lecular interactions of CO2 with the substrates. The sure, while the stronger band does not shift. For
FT-IR measurement was made for a single substrate methyl vinyl ketone, n
U
G
give several FT-IR absorption bands assignable to methyl acrylate, the n
(C=C) at 1600 cmÀ1 and d(CH) at 900–1000 cmÀ1. seen at around 1740 cmÀ1 with a very weak n
These absorption bands were not strong but the spec- absorption at 1670 cmÀ1. One may see a red-shift of
tra did not seem to change with CO2 pressure. the n(C=O) absorption peak with CO2 pressure but
Figure 8 presents FT-IR spectra in the range of 1800– not so significantly. From these FT-IR results it is im-
1600 cmÀ1, where the absorption bands of n
(C=O) possible to consider the electron negative or positive
and n
(C=C) appeared,[10] for the three dienophiles in features of the C-1 and C-2 carbon atoms of the three
ACHTREUNG
A
G
ACHTREUNG
AHCTREUNG
AHCTREUNG
ACHTREUNG
dense phase CO2 at a reaction temperature of 353 K. dienophiles; however, interactions of dense phase
The FT-IR absorption also occurred in different CO2 with the carbonyl group may be stronger for
ranges of wavenumber but the absorption bands were acrolein compared with the other two carbonyl com-
weak and these were not used for the present purpose pounds, as estimated from the peak shift of n(C=O)
G
Figure 8. FT-IR spectra of methyl acrylate (a), methyl vinyl ketone (b), and acrolein (c) dissolved in CO2 at different pres-
sures of 0.1, 4, 6, 8, 12, 16, 20 MPa (from bottom to top) and at 353 K. For a and b, the absorption at 8 MPa was weak and it
is increased by three and two times, respectively, herein.
1622
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1615 – 1625

Impact of Carbon Dioxide Pressurization on Liquid Phase Organic Reactions
FULL PAPERS
the 3 to 2 ratios among the three dienophiles may
result from steric effects because the 3 to 2 ratio is in
the opposite order to the electron withdrawing nature
À
À
À
CHO> COCH3 > COOCH3.
Similar explanations would be possible for the sol-
vent effects of ethanol and toluene on the product se-
lectivity (Table 1). The former HBD solvent can inter-
act with the carbonyl group through hydrogen bond-
ing and affect the product selectivity, similar to the
impact of the interaction with dense phase CO2 mole-
cules as discussed above. The use of ethanol increased
the rate of Diels–Alder reaction (Figure 7) compared
with toluene. However, the pressurization with CO2
caused the dilution of the reacting species, in addition
to the CO2-carbonyl interactions, and so it simply re-
sulted in a decrease of the reaction rate.
Scheme 1. Two approaches, a and b, yielding [2] adduct and
[3] adduct.
Conclusions
absorption band at a larger wavenumber with increas- In the present work we have investigated the Heck
ing CO2 pressure. The change in the product selectivi- reactions of methyl acrylate with various bromoar-
ty of the Diels–Alder reaction is most significant for enes in n-hexane pressurized by CO2 at pressures up
acrolein. Thus, CO2-carbonyl group molecular interac- to 14 MPa. The following conclusions are drawn with
tions would play a role in the alternation of the prod- respect to the impacts of the CO2 pressurization on
uct selectivity with increasing CO2 pressure (Table 1). the true liquid phase reactions. Indeed, the conversion
As described above, the CO2 molecules dissolved in is influenced by the CO2 pressurization although the
the liquid phases should have significant effects on reaction mixture does not include any gaseous reac-
the reactions. One is a simple dilution effect wheerby tants. The effects of dense phase CO2 strongly depend
the concentrations of the reacting species are reduced on the structures of bromoarenes used. The conver-
by the dissolution of CO2 and the rate of reaction be- sion increases with CO2 pressure for either 2-bromo-
comes decreased at increasing CO2 pressure
A
(Figure 7). The change of the 3 to 2 ratio with CO2 imum at a medium pressure of 2–4 MPa. However,
pressure may be ascribable to another effect of modi- such promotion does not occur with the other 3- and
fication of reactivity of reacting species through inter- 4-isomers of these substrates. The CO2 pressurization
actions of CO2 molecules in the liquid phases. The au- increases the volume of the liquid reaction phase due
thors speculate that CO2 molecules interact with the to the dissolution of CO2 molecules, which causes a
carbonyl group of a dienophile and this makes its C-1 decrease in the concentration of the reacting species.
carbon more positive compared to the C-2 carbon. As This negative effect on the rate of Heck reaction may
a result, the approach b of Scheme 1 may be more fa- be overthrown by a positive effect with CO2 dissolved
vorable to occur than that between the dienophile in the liquid reaction phase for these two bromoar-
and isoprene in which C-3 may be more positive than enes. The high pressure FT-IR results suggest interac-
C-2 owing to the electron-withdrawing nature of the tions of the bromarenes with the CO2 molecules,
methyl group attached to C-2. Thus, the 3 to 2 ratio which might play a role for the rate enhancement ob-
may be increasing with the CO2 pressure. As another served. Unfortunately, such an enhancement does not
possibility it may also be suggested that the interac- appear with the other substrates such as bromoben-
tions of CO2 molecules with the carbonyl group cause zaldehydes and methyl trans-bromocinnamates. No
steric effects on the product selectivity. We envisage promotion effect of CO2 pressurization was observed
that a large clump of the carbonyl group surrounded for liquid-solid Diels–Alder reactions of isoprene with
by CO2 molecules is formed through the interactions. methyl acrylate, methyl vinyl ketone, and acrolein
This may make the approach b of Scheme 1 more using a heterogeneous SiO2·Al2O3 catalyst in toluene
likely to occur compared with approach a, resulting in and ethanol. It should be noted, however, that the [3]
the increase of the 3 to 2 ratio with the CO2 pressure. and [2] adduct ratio does change with the CO2 pres-
These effects should be most significant for acrolein, sure. The most significant change occurs for acrolein
of which the carbonyl group could have most signifi- and the FT-IR spectroscopy measurements suggest
cant interactions with CO2, compared with methyl that its reactivity is altered through interactions with
vinyl ketone and methyl acrylate. The differences in CO2 molecules under pressurized conditions.
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1623

S. Fujita et al.
FULL PAPERS
with a triglycine sulfate (TGS) detector at a wavenumber
resolution of 2 cmÀ1 by using the absorption spectrum of
pure CO2 at the same pressure as background.
Experimental Section
Heck and Diels–Alder Reactions
A stainless steel autoclave reactor (50 cm3) was used to run
Heck reactions in an organic solvent (toluene) pressurized
with CO2 at different pressures up to 13 MPa. Typical reac-
tion conditions were: palladium acetate (Wako) 0.1 mmol,
triphenylphosphine (Wako) 0.4 mmol, methyl acrylate (Al-
drich) 10 mmol, aryl halide 10 mmol, triethylamine (Wako)
10 mmol, toluene (Wako) 10 cm3. All the chemicals were
used as received. The reactor was charged with these quanti-
ties of substrates, base, catalyst, and toluene and purged
with CO2 three times. The reactor was then heated in an oil
bath while mixing the reaction mixture by a Teflon-coated
magnetic stirrer. When the reactor temperature reached to
the desired value of 393 K, CO2 was introduced to the de-
sired pressure by a backpressure regulator (JASCO SCF-
Bpg) and the reaction time was measured. Then, the reactor
was cooled down to room temperature and depressurized
carefully. The reaction products were analyzed by gas chro-
matography (GL Science GC-390B) using iodobenzene as
an internal standard and gas chromatography-mass spec-
trometry (Shimadzu GC-MS QP5050 A). The conversions
were determined from the amounts of aryl halides reacted.
The same high pressure reaction system was also used to
run Diels–Alder reactions of isoprene with the different di-
enophiles methyl acrylate, methyl vinyl ketone, and acrolein.
The reaction procedures were the same as used for the
Heck coupling runs and for the GCanalysis, decane was
used as an internal standard. A heterogeneous silica-alumi-
na, JRC-SAH-1 supplied by Catalysis Society of Japan, was
used as a catalyst. Under the experimental conditions used,
the total conversion and the product yield were reproducible
in relative errors of within Æ5%.
Acknowledgements
This work was supported by Japan Society for the Promotion
of Science with Grant-in-Aid for Scientific Research (B)
18360378.
References
[1] a) G. Musie, M. Wei, B. Subramaniam, D. H. Busch,
Coord. Chem. Rev. 2001, 219–221, 789; b) M. Wei, G. T.
Musie, D. H. Busch, B. Subramaniam, J. Am. Chem.
Soc. 2002, 124, 2513; c) B. Subramaniam, C. J. Lyon, V.
Arunajatesan, Appl. Catal. B Environ. 2002, 37, 279;
d) B. Kerler, R. E. Robinson, A. S. Borovik, B. Subra-
maniam, Appl. Catal. B Environ. 2004, 49, 91; e) M.
Wei, G. T. Musie, D. H. Busch, B. Subramaniam, Green
Chem. 2004, 6, 387.
[2] a) G. Combes, E. Coen, F. Dehghani, N. Foster, J. Su-
percrit. Fluids, 2005, 36, 127; b) X. Xie, C. L. Liotta,
C. A. Eckert, Ind. Eng. Chem. Res. 2004, 43, 7907;
c) H.-S. Phiong, E. P. Lucien, A. A. Adesina, J. Super-
crit. Fluids, 2003, 25, 155; d) D. Chouchi, D. Gourgouil-
lon, M. Courel, J. Vital, M. N. da Ponte, Ind. Eng.
Chem. Res. 2001, 40, 2551.
[3] a) H. Jin, B. Subramaniam, A. Ghosh, J. Tung, AIChE
J. , 2006, 52, 2575; b) H. Jin, B. Subramaniam, Chem.
Eng. Sci., 2004, 59, 4887; c) O. Hemminger, A. Marteel,
M. R. Mason, J. A. Davies, A. R. Tadd, M. A. Abra-
ham, Green Chem. 2002, 4, 507.
[4] a) F. Zhao, S. Fujita, J. Sun, Y. Ikushima, M. Arai,
Chem. Commun. 2004, 2326; b) F. Zhao, S. Fujita, S.
Akihara, M. Arai, J. Phys. Chem. A 2005, 109, 4419;
c) R. Liu, F. Zhao, S. Fujita, M. Arai, Appl. Catal. A
Gen. 2007, 316, 127.
[5] a) V. Farina, Adv. Synth. Catal. 2004, 346, 1553; b) P.
Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100,
3009; c) B. Cornils, W. A. Herrmann, (Eds.), Applied
Homogeneous Catalysis with Organometallic Com-
pounds, Wiley-VCH, Weinheim, 2000; d) A. M. Trze-
ciak, J. J. Ziolkowski, Coord. Chem. Rev. 2007, 251,
1281.
[6] a) O. Reiser, Top. Catal. 1998, 5, 105; b) K. Matsumoto,
A. Sera, T. Uchida, Synthesis 1985, 1; c) K. Matsumoto,
A. Sera, Synthesis 1985, 999; d) F. Diederich, P. J.
Stang, (Eds.), Metal-catalyzed Cross-coupling Reactions,
Wiley-VCH, Weinheim, 1998, pp 111–166; e) R. van
Eldik, F.-G. Klaerner, (Eds.), High Pressure Chemistry,
Wiley-VCH, Weinheim, 2002, pp 230–238.
[7] a) R. D. Weinstein, A. R. Renslo, R. L. Danheiser, J. W.
Tester, J. Phys. Chem. B 1999, 103, 2878; b) Y. Ikushi-
ma, M. Arai, in: Chemical Synthesis Using Supercritical
Fluids, (Eds.: P. G. Jessop, W. Leitner), Wiley-VCH,
Weinheim, 1999, pp 262–265; c) E. J. Beckman, J. Su-
percrit. Fluids, 2004, 28, 121; d) D. Clarke, M. A. Ali,
A. A. Clifford, A. Parratt, P. Rose, D. Schwinn, W.
Phase Behavior and FT-IR Measurements
The phase behavior was examined by visual observation
using an autoclave with a transparent glass window. A de-
sired volume of a toluene phase was added to the autoclave
with the volume ratio of toluene/autoclave similar to that
used in Heck reaction runs. An atmospheric pressure of
CO2 was introduced and the autoclave was closed and
heated to the desired reaction temperature of 393 K. While
keeping the autoclave at this temperature, more CO2 was in-
troduced up to higher pressures. The phase behavior was
visually examined at different CO2 pressures.
The FT-IR spectra of selected substrates in dense phase
CO2 were measured with an in situ high pressure FT-IR
spectrometer, JASCO FTIR-620, equipped with a 1.5 cm3
high pressure cell using crystalline ZnS windows with an op-
tical path length of 4 mm.[4] The FT-IR measurements were
made at 393 K and at CO2 pressures up to 12 MPa. A small
volume (about 0.01 cm3) of a sample carbonyl compound
was introduced into the cell and it was purged with CO2
three times. The cell was heated up to a temperature of
393 K by the circulation of preheated oil around it
(JULABO F25-HP). Then, the pressure was increased
slowly by introduction of liquid CO2 by a pressure pump
while stirring by a Teflon coated magnetic stirrer. When the
pressure reached a certain value, the stirring was continued
for 2 min and stopped. The FT-IR spectra were measured
1624
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1615 – 1625

Impact of Carbon Dioxide Pressurization on Liquid Phase Organic Reactions
FULL PAPERS
Bannwarth, C. M. Rayner, Curr. Top. Medic. Chem.
2004, 4, 729; e) F. Fringuelli, A. Taticchi, The Diels–
Alder Reaction – Selected Practical Methods, Wiley,
West Sussex, 2002.
(Eds.), Chemical Synthesis Using Supercritical Fluids,
Wiley-VCH, Weinheim, 1999.
[14] a) Y. Ikushima, N. Saito, M. Arai, K. Arai, Bull. Chem.
Soc. Jpn. 1991, 64, 2224; b) Y. Ikushima, N. Saito, M.
Arai, J. Phys. Chem. 1992, 96, 2293; c) R. Eberhardt, S.
Loebbecke, B. Neidhart, C. Reichardt, Liebigs Ann./
Recueil, 1997, 1195; d) J. A. Hyatt, J. Org. Chem. 1984,
49, 5097.
[15] F. Zhao, B. M. Bhanage, M. Shirai, M. Arai, Chem.
Eur. J. 2000, 6, 843.
[16] a) K. Berlier, M. Frꢂre, J. Chem. Eng. Data 1997, 42,
533; b) M.-G. Olivier, R. Jadot, J. Chem. Eng. Data
1997, 42, 230.
[17] J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic
Chemistry, Oxford University, Oxford, 2001, p 917.
[18] a) T.-K. Ha, C. Pal, P. N. Ghosh, Spectrochim. Acta A
1992, 48, 1083; b) Y. Akiyama, S. Fujita, H. Senboku,
C. M. Rayner, S. A. Brough, M. Arai, J. Supercrit.
Fluids, in press (doi:10.1016/j.supflu.2008.03.009).
[8] a) F. Zhao, B. M. Bhanage, M. Shirai, M. Arai, Stud.
Surf. Sci. Catal. 1999, 122, 427; b) A. F. Schmidt, V. V.
Smirnov, A. Al-Halaiga, Kinet. Catal. 2007, 48, 390.
[9] J. C. Meredith, K. P. Johnston, J. M. Seminario, S. G.
Kazarian, C. A. Eckert, J. Phys. Chem. 1996, 100,
10837–10848.
[10] K. Nakanishi, P. H. Solomon, Infrared Absorption Spec-
troscopy, 2nd edn., Holden-Day, San Francisco, 1977.
[11] C. Reichardt, Solvents and Solvent Effects in Organic
Chemistry, 3rd edn., Wiley-VCH, Weinheim, 2003.
[12] M. Khajehpour, J. F. Kauffman, J. Phys. Chem. A 2000,
104, 9512.
[13] a) T. Moriyoshi, T. Kita, Y. Uosaki, Ber. Bunsenges.
Phys. Chem. 1993, 97, 589; b) P. G. Jessop, W. Leitner,
Adv. Synth. Catal. 2008, 350, 1615 – 1625
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1625