The improved Kolbe-Schmitt reaction using supercritical carbon dioxide

Article abstract of DOI:10.1016/j.tetlet.2007.05.132

A comparative analysis of the reactions between sodium phenoxide and CO₂ under the gaseous and supercritical conditions has demonstrated the preferential effect of supercritical conditions for promoting the reactivity of CO₂. A significant increase in product yield was obtained in supercritical CO₂ compared to reactions undertaken in gaseous CO₂.

Full text of DOI:10.1016/j.tetlet.2007.05.132

Tetrahedron Letters 48 (2007) 5309–5311
The improved Kolbe–Schmitt reaction using supercritical
carbon dioxide
Takayuki Iijima and Tatsuaki Yamaguchi*
Department of Life and Environmental Sciences, Faculty of Engineering, Chiba Institute of Technology, 2-17-1 Tsudanuma,
Narashino, Chiba 275-0016, Japan
Received 14 March 2007; revised 16 May 2007; accepted 17 May 2007
Available online 26 May 2007
Abstract—A comparative analysis of the reactions between sodium phenoxide and CO2 under the gaseous and supercritical condi-
tions has demonstrated the preferential effect of supercritical conditions for promoting the reactivity of CO2. A significant increase
in product yield was obtained in supercritical CO2 compared to reactions undertaken in gaseous CO2.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The conversion of carbon dioxide into industrially use-
ful compounds has recently attracted much interest in
view of the so-called ‘Green Chemistry’ and ‘Sustainable
Society’ concepts. As a potential pathway for the effec-
tive utilization of carbon dioxide as a safe and cheap
C1 building block in organic synthesis,1–3 the synthesis
of hydroxybenzoic acid via coupling reactions between
carbon dioxide and phenols has attracted much atten-
tion. This is because hydroxybenzoic acid can be used
in many application areas as valuable feedstock and as
an intermediate in the production of pharmaceuticals
and fine chemicals. However, the conventional Kolbe–
Schmitt reaction4 (Scheme 1), which is a heterogeneous
reaction in the gas–solid phase, requires a high temper-
ature and long reaction time. The yields of aromatic
hydroxycarboxylic acids in polar aprotic solvents were
much greater than conventional Kolbe–Schmitt reac-
tion. The polar aprotic solvents aid in the dissolution
of the metal salt, forming a homogeneous solution. This
is because polar solvents solvate the cation–anion pair
by increasing their intramolecular distances.5,6
Supercritical fluid (SCF) medium has recently attracted
much attention as the fourth phase for performing
chemical synthesis, besides gas, liquid, and solid phases.
The physical properties of SCFs can be continuously
varied from those analogous to gases to those analogous
to liquids by manipulating its pressure and/or tempera-
ture; hence, SCFs are expected to enhance the reaction
rate and product selectivity in comparison with conven-
tional organic solvents.7–9 Among SCFs, supercritical
carbon dioxide is the most attractive reaction medium
because of its low critical temperature (304 K) and pres-
sure (7.3 MPa), as well as due to its nontoxic and non-
flammable nature and low cost.
Here we wish to report the carboxylation of sodium
phenoxide10 promoted by supercritical carbon dioxide.
First, to compare supercritical CO2 with gaseous
CO2, we systematically investigated the reaction under
a number of pressures ranging from 3 to 17 MPa.11
Figure 1 presents the effect of CO2 pressure on the con-
version of sodium phenoxide to hydroxybenzoic acid.
The yield drastically changes with variations in the
CO2 pressure demonstrating the preferential effect of
the supercritical conditions in promoting the reactivity
of CO2 as seen in Figure 1. The conversion of sodium
ONa
OH
OH
CO2
COONa
+
COONa
Scheme 1. Kolbe–Schmitt reaction of sodium phenoxide.
Keywords: Supercritical carbon dioxide; Sodium phenoxide; Salicylic
acid; p-Hydroxybenzoic acid; Kolbe–Schmitt reaction.
CAUTION! Experiments, using compressed gases such as supercri-
tical CO2, are potentially hazardous and must only be carried out
using appropriate equipment and safety precautions.
*
Corresponding author. Tel./fax: +81 47 478 0420; e-mail: tatsuaki.
yamaguchi@it-chiba.ac.jp
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.05.132

5310
T. Iijima, T. Yamaguchi / Tetrahedron Letters 48 (2007) 5309–5311
100
80
60
40
20
0
10
tion in a pure CO2 solvent. The decrease in the conver-
sion with an increased N2 presence in the reaction fluid
is very pronounced at the two reaction pressures, but the
product selectivity was unaffected, as can be seen in Fig-
ure 3. There is no doubt that the presence of N2 in the
supercritical region decreases the conversion of phen-
oxide. Zhang and King15 reported that the density of
the fluid is linearly proportional to the dissolved mixture
content in the CO2. However, this relationship is not lin-
ear. This result suggests the density of the fluid alone
does not explain the conversion drop in CO2–N2 mix-
tures. The significant drop of in the conversion of
sodium phenoxide in binary mixtures may involve a
disruption of the CO2 shell around the phenoxide.
8
6
4
2
0
Pc
0
5
10
15
20
In summary, we have reported that a step in developing
an efficient and mild method for the synthesis of
hydroxybenzoic acid by using supercritical CO2, an ideal
PCO / MPa
2
Figure 1. Effect of CO2 pressure on the conversion of sodium
phenoxide. Reaction conditions are as follows: a 50-mL stainless steel
autoclave; sodium phenoxide, 10 mmol; reaction time, 1 h; reaction
temperature, 393 K; (s) conversion, (n) o-/p-ratio.
100
80
phenoxide tends to be dependent on the CO2 pressure at
lower pressures, but is remarkably lower than that under
supercritical conditions. This phenomenon may be
explained by a local augmentation effect;12 the reaction
rate increases because of the clustering of solvent mole-
cules around the solute species.13 The apparent increase
in the conversion for the reaction with CO2 is likely due
to the increased local concentration of CO2 around the
sodium phenoxide. Sodium phenoxide is dissolved in
supercritical CO2, creating a homogeneous solution.
This is done because the carboxylation of naphthoxide
readily occurs through the use of a polar solvent. How-
ever, the reaction system seems to have two phases due
to the low solubility of sodium phenoxide in supercriti-
cal CO2. The reason is that supercritical CO2 has a low
dielectric constant, with a value generally between 1.1
and 1.5. Furthermore, the conversion decreases with
an increase in the CO2 pressure above 10 MPa, because
the local solvation occurs in supercritical fluids at condi-
tions near the critical pressure. Interestingly, in the CO2
pressure range from 8 to 17 MPa, we found that the
product selectivity is almost independent providing p-
hydroxybenzoic acid with 20% selectivity. In the low-
pressure region, the reaction proceeds with the forma-
tion of a sodium phenoxide–CO2 complex as an interme-
diate. On the other hand, in the high-pressure region,
the reaction occurs by the dissolution of phenoxide in
CO2. The difference in the reaction mechanism in the
low- and high-pressure regions results in some change
in product selectivity.
60
40
20
0
0.0
0.2
0.4
0.6
0.8
1.0
PCO2 P-1 / -
Figure 2. Sodium phenoxide conversion in CO2–N2 having different
N2 content and at selected pressures. Reaction conditions are as
follows: Sodium phenoxide, 10 mmol; reaction time: 1 h; reaction
temperature: 393 K; (s) 8 MPa, (d) 15 MPa.
100
80
60
40
20
We conducted similar experiments with a variety of
compositions in the supercritical region in order to
determine whether the density effect is the principle
cause for the rate enhancement effect. Because melting
is normally initiated by the formation of a liquid in
CO2–N2 inclusions,14 N2 was selected for the entrainer.
The dependence of the conversion on the fluid composi-
tion of the CO2–N2 system at 8 and 15 MPa is shown in
Figure 2. The right ends of the plots represent the reac-
0
0.0
0.2
0.4
0.6
0.8
1.0
PCO P-1 / -
2
Figure 3. A plot of product selectivity from carboxylation of sodium
phenoxide in CO2–N2 having different N2 content and at selected
pressures. Reaction conditions are as follows: Sodium phenoxide,
10 mmol; reaction time: 1 h; reaction temperature: 393 K; (n) oHBA-
8 MPa, (m) pHBA-8 MPa, (s) oHBA-15 MPa (d) pHBA-15 MPa.

T. Iijima, T. Yamaguchi / Tetrahedron Letters 48 (2007) 5309–5311
5311
8. Hyde, J. R.; Licence, P.; Carter, D. N.; Poliakoff, M. Appl.
Catal., A: Gen. 2001, 222, 119–131.
9. Beckman, E. J. J. Supercrit. Fluids 2004, 28, 121–
191.
10. Preparation of sodium phenoxide: Phenol (9.4 g, 100 mmol)
was dissolved in 100 mL of an aqueous solution of sodium
hydroxide, followed by stirring at room temperature for
1 h. The resultant solution was placed in a rotary
evaporator and heated at 353 K for 3 h. The resulting
solid was dissolved in a large amount of ethanol and
heated at 333 K for 3 h. The white solid thus obtained was
further dried in vacuo at 473 K for 6 h to obtain pure
sodium phenoxide.
reaction medium in view of green chemistry, as a reac-
tant and a reaction medium. The reaction in supercriti-
cal CO2 proceeds dramatically as compared with the
conventional Kolbe–Schmitt reaction, presumably due
to the increase in the density of CO2 around the phenox-
ide. The addition of nitrogen as an entrainer led to a
decrease in the reaction rate through the disintegration
of the CO2 shell in the vicinity of sodium phenoxide.
Further work may include trying this reaction at even
higher temperatures to further enhance the reaction rate.
11. General procedure and analysis: The reaction was per-
formed with a supercritical CO2 reactor system including
SCF-Get, SCF-Bpg, and SCF-Sro (JASCO). The prepared
sodium phenoxide was placed in a 50-mL stainless steel
autoclave (SUS-316) containing a stirring bar. Liquid
CO2, cooled to 263 K, was introduced into the autoclave,
which had been heated to 393 K, using an high perfor-
mance liquid chromatography (HPLC) pump after flush-
ing with nitrogen. The resultant mixture was stirred for 1 h
at 393 K with the CO2 pressure constantly maintained at
the requisite pressure. The autoclave was cooled in an ice
bath, and the pressurized CO2 was gradually released. The
resulting solid mixture was washed from the autoclave
with deionized water. The solution was analyzed by HPLC
with an ODS column (250 · 4.6 B mm) on a Shimadzu
LC-10AD chromatograph.
References and notes
1. Arakawa, H.; Areata, M.; Armor, J. N.; Barteau, M. A.;
Beckman, E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.;
Dinjus, E.; Dixon, D. A.; Domen, K.; DuBois, D. L.;
Eckert, J.; Fujita, E.; Gibson, D. H.; Goddard, W. A.;
Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H. H.;
Manzer, J. E.; Marks, T. J.; Morokuma, K.; Nicholas, K.
M.; Periana, R.; Que, L.; Rostrup-Nielson, J.; Sachtler,
W. M. H.; Schmidt, L. D.; Sen, A.; Somorjai, G. A.; Stair,
P. C.; Stults, B. R.; Tumas, W. Chem. Rev. 2001, 101, 953–
996.
2. Cott, D. J.; Ziegler, K. J.; Owens, V. P.; Glennon, J. D.;
Graham, A. E.; Holtmes, J. D. Green Chem. 2005, 7, 105–
110.
12. Brennecke, J. F.; Chateauneuf, J. E. Chem. Rev. 1999, 99,
433–452.
3. Hyde, J. R.; Walsh, B.; Singh, J.; Poliakoff, M. Green
Chem. 2005, 7, 357–361.
13. Oakes, R. A.; Clifford, A. A.; Rayner, C. M. J. Chem.
Soc., Perkin Trans. 1 2001, 917–941.
14. van den Kerkhof, A. M. Geochim. Cosmochim. Acta 1990,
54, 621–629.
4. Lindsey, A. S.; Jeskey, H. Chem. Rev. 1957, 57, 583–619.
5. Baxter, J.; Yamaguchi, T. J. Chem. Res. 1997, 374–375.
6. Iijima, T.; Iwase, T.; Yamaguchi, T. J. Jpn. Petrol. Inst.
2006, 49, 206–209.
15. Zhang, Z.; King, J. W. J. Chromatogr. Sci. 1997, 35, 483–
488.
7. Noyori, R., Ed.; Supercritical Fluids. Chem. Rev. 1999, 99,
353–389.