Aqueous emulsion containing fluorous cobalt species in supercritical CO2 for catalytic air oxidation of toluene

Article abstract of DOI:10.1039/b205673a

An aqueous emulsion containing ionic Co²⁺ and Br^- species stabilised by fluorous surfactant-like species in supercritical CO₂-air mixture acts as a nano-reactor with excellent interfacial contacts of all necessary hydrophi

Full text of DOI:10.1039/b205673a

Aqueous emulsion containing fluorous cobalt species in supercritical CO2
for catalytic air oxidation of toluene
Jie Zhu, Alan Robertson and Shik Chi Tsang*
Surface and Catalysis Research Centre, Department of Chemistry, University of Reading, Whiteknights,
Reading RG6 6AD, UK. E-mail: s.c.e.tsang@reading.ac.uk
Received (in Cambridge, UK) 11th June 2002, Accepted 30th July 2002
First published as an Advance Article on the web 16th August 2002
An aqueous emulsion containing ionic Co2+ and Br2 species
stabilised by fluorous surfactant-like species in supercritical
CO2–air mixture acts as a nano-reactor with excellent
interfacial contacts of all necessary hydrophilic/hydrophobic
species, which renders safe operation of catalytic aerial
oxidation of toluene at high yields.
quencies (TOF = 6.19 3 1023 s21) were achieved without the
use of acetic acid. The nature of the emulsion droplet (nano-
reactor) containing Co2+/NaBr (allowing excellent interfacial
contacts between non-polar substrate molecules and ionic
catalytic species) coupled with the intrinsic advantages of
supercritical CO2 [low viscosity, high diffusivity, rapid product
desorption, and complete miscibility of substrates (alkylaro-
matics and dioxygen in particular)] enables a fast oxidation
reaction to be sustained.
Selective catalytic partial oxidation of alkylaromatics to acids
by air or molecular oxygen is of great economic and industrial
importance.1 The key challenges regarding the development of
new catalytic processes in this area include the identification of
active and particularly selective catalyst species (unselective
attacks on C–C or aromatic C–H bonds should be avoided), ease
of separation of product and catalyst and safe operation of the
highly exothermic oxidation reactions. Many attempts have
been made at developing solid catalysts containing transition
metal ions. However, poor activity and selectivity and problems
with catalyst stability presently excludes them from existing
applications. Currently, the principal industrial methods employ
simple cobalt salts as homogeneous catalysts in an air
pressurised aqueous–acetic acid mixture in the presence of ionic
promoter(s) (Mn ions and/or bromide).2 The acetic acid–water
mixture is apparently crucial in providing the solvent medium
for the free and effective interactions between ionic catalyst and
promoter species and the organic substrates. However, prob-
lems with decarboxylation of the solvent and products due to the
unselective oxidation, use of extremely corrosive and toxic
acetic acid, difficulty in solvent and catalyst recovery, and
explosion hazards associated with solvent and dioxygen at high
concentration (with reported explosions) are obvious.3 With
increasingly demanding environmental legislation, public and
corporate pressure, a new process is urgently required.4
Supercritical carbon dioxide (scCO2) has recently received
considerable attention as a new versatile, environmentally
friendly medium for a variety of catalytic reactions; these
include free-radical polyerization,5 hydroformylation,6 hydro-
genation,7 and partial oxidations.8 Its non-toxic, non-combusti-
ble and non-flammabile nature combined with the ease of
solvation of some organic substrates in the inexpensive medium
and ease of pressure facilitated separation may offer an exciting
possibility of replacing acetic acid for catalytic partial oxida-
tions. Despite these encouraging facts, the major limitation is its
inability to dissolve hydrophiles, in particular ionic catalyst/
promoter species that are required for the oxidation (Co2+/
NaBr). Recent work has shown that a number of fluorous
surfactants were able to form aqueous microemulsions in
scCO2.9 Similarly, other fluorous related species can also
modify the interfacial tension of the scCO2 forming a
metastable aqueous emulsion therein.10 Thus, fluorous surfac-
tants and related compounds are able to disperse water-soluble
metal salts in the nonpolar scCO2 medium.
Experiments were carried out in a 160 ml stainless steel
autoclave (with Teflon cup insert reducing the volume to ca.
111 ml) equipped with an overhead stirrer. 0.25 mmol
[CF3(CF2)8COO]2Co·nH2O denoted as F-Co,† 0.2 mmol NaBr
and a small quantity of H2O (100 ml unless otherwise specified)
were added to the autoclave that hosted a small internal
container holding 18.8 mmol toluene liquid (in order to avoid its
direct liquid-phase reaction with the catalyst). The autoclave
was charged with 10 bar O2 and topped up to 150 bar by
pumping CO2 into the autoclave at 120 °C. The amount of
toluene used was readily soluble under these conditions.11 The
autoclave was connected to a GC for on-line monitoring of
reaction mixtures intermittently. All products were analyzed
quantitatively using GC and HPLC after 12 h.
The non-fluorous cobalt catalyst (Co(II) acetate/NaBr) that is
known to be extremely active in the oxygen pressured acetic
acid–water system, was found to be inert (conv. 0.1%) due to its
insolubility in CO2–water (monitored through sapphire win-
dows). In contrast, F-Co/NaBr showed excellent activity (conv.
98.2% with selectivity to benzoic acid of 99.1%). It was
interesting to find that the F-Co showed a relatively poor
activity when either NaBr or water (using dehydrated F-Co) was
not included (conv. 6.0 and 14.2%, respectively). These results
clearly indicate that F-Co, NaBr and H2O are all essential
components for effective catalysis. Variation of the quantity of
water added (100, 200, 300, 500 ml) to the F-Co/NaBr was then
investigated (Fig. 1). The activity curves all show a general
induction period followed by a rapid increase in activity. It is
interesting that the induction period was shortened in ac-
cordance with increasing quantity of water added. It is well
documented that formation of emulsion assemblies depend on
the quantity and ratio of water to surfactant used (W) in organic
solvent. It is also known that scCO2 can dissolve water in its
phase. Some water is therefore needed to saturate the scCO2
phase before emulsion droplets can form.9a Thus, our results
suggest that emulsions may not form straight away if in-
sufficient amounts of water are present during the induction
period (poor solid/CO2 interface due to lack of water).
However, oxidation of toluene will yield additional quantities of
Here we report a new approach for oxidation of toluene by
dioxygen based on emulsion catalysis in supercritical CO2 using
fluorous-tag Co(II)/NaBr in a water–scCO2 mixture (the
fluorous tag and scCO2 solvent contain no C–H for possible
non-selective attacks). Extremely high conversion ( > 99%),
high selectivity to benzoic acid ( > 99% with traces of
benzaldehyde and benzyl alcohol), and high turnover fre-
Fig. 1 The effect of water on catalytic activity.
2044
CHEM. COMMUN., 2002, 2044–2045
This journal is © The Royal Society of Chemistry 2002

View Online
water (by-product) to assist emulsion formation at a later stage.
It is not yet known whether a truly stable microemulsion or a
coarsely dispersed biphasic system is involved. However, as far
as catalysis is concerned, once the emulsion is established, it can
account for the abrupt increase in activity after the induction
period. Based on the activity curves, we estimate that a
minimum of 270–310 ml water (the amount of water added plus
the envisaged amount of water produced from reaction at the
end of the induction period) is required before the emulsion is
formed. Despite the actual water solubility depending on
conditions and impurities (substrates and products) in scCO2
this value is in a good agreement with Wiebe’s calculations.12
The corrected Wo values of water to fluorous Co(II) ratios after
deducing the amount water required to saturate the scCO2 from
0–44. The effects of PO , toluene, NaBr, PCO (total P), and
Fig. 2 The effect of cation on catalytic activity.
However, it can be postulated that interactions of the fluorous
anions with scCO2/water to form emulsions may be kinetically
slower in order to override the stronger ionic interactions within
this solid (reflected by its higher mp). The extremely small size
of the emulsion droplets in scCO2 are reflected in a detailed
postmortem analysis of the catalyst following a rapid de-
pressurisation of the system (F-Co(II)/NaBr/400 ml H2O) from
typical conditions to give a solid foam. TEM micrographs of
this foam (not shown) indicate extremely well dispersed NaBr
2
2
temperature on activity were briefly studied. From Table 1, a
critical quantity of NaBr seems to be required ( > 0.05 mmol) for
high activity. Once above this value the activity remains
virtually the same. This is consistent with the literature where;
a small quantity of Br2 acts as ionic promoter to facilitate rapid
electron transfers of the catalytic Co species (II and III). The
activity observed follows a first order rate with respect to
toluene added suggesting the C–H activation in the toluene
oxidation is likely be the rate limiting step as is typical in most
crystallites of 3 ± 1 nm in the non-crystalline fluorous Co(II
)
matrices. Such small but uniform size NaBr crystallites with
excellent dispersion in F-Co(II) reveals the high dispersity of the
emulsion in scCO2. It is noted that Zielinski et al.13 showed
micelle sizes in scCO2 of about 4.0–7.2 nm. To conclude,
water–scCO2 is shown to be an excellent alternative solvent to
acetic acid for the important Co(II) air-oxidation process. The
generic concept of using H2O–scCO2 emulsions to bring
species of different polarities into contact with excellent mass
and heat transfers in sustaining a fast catalytic reaction is now,
for the first time, demonstrated. We believe that this novel nano-
reactor system could be utilized for a wide range of oxida-
tions.
alkane oxidations.1 Activity was found to be independent of PO
2
(23.1–46.2 mmol, rendering O2 in excess), which reflects ease
of availability of O2 for the catalytic sites (miscible in scCO2
with excellent diffusivity) in scCO2. It is also interesting that a
higher rate was achieved with lower PCO (toluene confirmed to
be solublized at all conditions), which2 may be attributed to
higher diffusivity and/or a greater concentration of toluene at
lower applied pressures. The excellent TOFs were recorded at
80 °C (9.15 3 1024 s21), 100 °C (2.86 3 1023 s21) and 120 °C
(6.48 3 1023 s21) and the apparent activation energy of this
catalysed oxidation reaction by this new system was estimated
to be 49.82 kJ mol21. Our un-optimized catalyst, evaluated at
100 °C (2.86 3 1023 s21) gave an activity for toluene oxidation
which was more than 10 times higher than the reported rate of
2.4 3 1024 s21 (87 °C) in acetic acid–water2a and 100 times
higher than the solid Co counterpart in scCO2 (1 3 1025 s21 at
140–200 °C11). Although the differences of these rates should
not be taken too literally, the excellent activity of our emulsion
catalyst in water–scCO2 is apparent. Such high activity can be
attributed to the intrinsically faster transfers in water–scCO2
than in liquid, the dynamic properties of emulsion and the local
high concentration of ingredients in the emulsion droplets.
Importance of the fluorous anion for aqueous emulsion
formation is clearly demonstrated in Fig. 2.The F-Mg (replacing
F-Co) blended with Co acetate (0.25 mmol) plot shows similar
activity to the F-Co. This result suggests that cationic species
are likely to be mobile (within the droplet) despite their
difference in binding with the fluorous anion (compare mp
values). It is therefore unnecessary to directly fluorinate the Co
species as the inexpensive cobalt acetate is equally effective
when in combination with fluorous anions. F-K/Co acetate
achieved comparable conversions though over a greater period
of time (Fig. 2). The precise reason for this is not yet known.
Notes and references
† Fluorous species synthesis, properties and characterisations: F-Co was
synthesised via addition of 4 mmol nonadecafluorodecanoic acid (NDFDA)
into 5 ml of a freshly prepared solution of CoCO3 (4 mmol) in methanol.
The mixture was kept at 40 °C and continually stirred. 10 ml diethyl ether
was then added to stabilise the fluorinated Co(II) species formed. As a result,
most of the solid was dissolved after 16 h. After filtering, a pink solid was
obtained following removal of solvent. This solid was dissolved in diethyl
ether and passed through a column of silica for purification. Chemical
microanalysis and atomic absorption showed the dried blue compound (pink
when hydrated) to be (CF3(CF2)8CO2)1.95Co. IR (Nujol) showed a strong
CNO absorption at 1657 cm21 (sharp) as in the carboxylic anion form while
the NDFDA has weaker absorption at 1717 (broad) in the acid form. This
fluorous Co salt was able to form a cloudy colloidal suspension in water
with stirring. It’s melting point was determined to be 138–139 °C. The
analogues of fluorous tagged Mg (mp 152–154 °C) and K (mp 205–207 °C)
were also synthesized.
1 A. K. Suresh, Ind. Eng. Chem. Res., 2000, 39, 3958.
2 (a) M. P. Czytko and G. K. Bub, Ind. Eng. Chem. Prod. Rev. De., 1981,
20, 481; (b) W. Partenheimer, Chem. Ind., 1990, 40, 321.
3 E. Sahle-Demessie, M. A. Gonzalez, J. Enriquez and Q. Zhao, Ind. Eng.
Chem. Res., 2000, 39, 4858.
4 S. A. Chavan, S. B. Halligudi, D. Srinivas and P. Ratnasmy, J. Mol.
Catal., 2000, 161, 49.
5 J. M. Desimone, E. E. Maury, Y. Z. Meneceloglu, J. B. Mcclain, T. J.
Romack and J. R. Combes, Science, 1994, 265, 356.
Table 1 Effect of experimental parameters on catalytic activity
6 R. J. Klingler and J. W. Rathke, J. Am. Chem. Soc., 1994, 116, 4772.
7 M. G. Hitzler and M. Poliakoff, Chem. Commun., 1997, 17, 1667.
8 A. M. Steele, J. Zhu and S. C. Tsang, Catal. Lett., 2000, 73(1), 9.
9 (a) K. P. Johnston, K. L. Harrison, M. J. Clarke, S. M. Howdle, M. P.
Heitz, F. V. Bright, C. Carlier and T. W. Randolph, Science, 1996, 271,
624; (b) D. E. Fremgen, E. S. Eugene, R. E. Gerald, R. J. Klingler and
J. W. Rathke, J. Supercrit. Fluid., 2001, 19, 287.
TOF
(3 103 s21
TOF
(3103 s21
)
Variablea
)
Variablea
PO /bar
5
7.5
10
0.2
0.5
0.8
1.5
3.94
3.94
3.96
0.50
1.40
1.98
3.96
NaBr/mmol
2
5
10
20
1.10
3.40
3.97
3.96
6.30
3.96
3.33
2
10 P. A. Psathas, E. A. Sander, M. Y. Lee, K. T. Lim and K. P. Johnston,
J. Dispersion Sci. Technol., 2002, 23, 65.
Toluene/mmol
Total P/bar 100
11 A maximum of 32.6 mmol toluene can be dissolved in ca. 111 ml scCO2
at 150 bar at 120 °C according to GC analysis using trace benzene as a
standard. Results are consistent with lit. solubility diagrams: K. M.
Dooley and F. C. Knopf, Ind. Eng. Chem. Res., 1987, 26, 1910.
12 R. Wiebe, Chem. Rev., 1941, 29, 475.
150
170
a Normal conditions: 10 bar O2, topped up with scCO2 to 150 bar, NaBr (0.2
mmol), toluene (14.1 mmol), H2O (400 ml) and F-Co (0.25 mmol) at 120 °C,
only one variable specified above was altered.
13 R. G. Zielinski, S. R. Kline, E. W. Kaler and N. Rosov, Langmuir, 1997,
13, 3934.
CHEM. COMMUN., 2002, 2044–2045
2045