Applied Catalysis A: General 298 (2006) 109–114

www.elsevier.com/locate/apcata

Catalytic oxidation of VOCs and CO by ozone over

alumina supported cobalt oxide
P. Konova a,*, M. Stoyanova b, A. Naydenov a, St. Christoskova b, D. Mehandjiev a
a
Institute of General and Inorganic Chemistry, Bulgarian Academy of Sciences,
Acad. G. Bonchev St., Bl. 11, 1113 Sofia, Bulgaria
b
Chemical Faculty, Paisii Hilendarski University of Plovdiv, Bulgaria
Received 16 July 2005; received in revised form 19 September 2005; accepted 23 September 2005
Available online 9 November 2005

Abstract
An alumina-supported cobalt oxide system with overstoichiometric oxygen (CoOx/Al2O3) was investigated with respect to heterogeneous
catalytic decomposition of ozone, complete oxidation of volatile organic compounds (VOCs) and oxidation of CO. The experiments were
performed in the temperature range of
45 to 250 8C in an isothermal plug-flow reactor. The characterization of the CoOx/Al2O3 catalyst was
performed by chemical analysis, XPS, XRD, IR techniques, magnetic and adsorption measurements.
A very high activity of the catalyst towards ozone decomposition was observed even at temperatures below
40 8C, the catalyst remaining
active for a long time. The activity of the catalyst with respect to complete oxidation of VOCs and oxidation of carbon monoxide was studied in
presence of different oxidizing agents (ozone or oxygen). A significant increase in the catalytic activity and decrease in the reaction temperature
was observed using ozone as an oxidant.
Two main reasons for this behaviour were found: (i) the high content of active and mobile oxygen obtained during the synthesis on the catalyst
surface; (ii) the catalytically active complex of O
[Co4+], formed during the reaction of ozone decomposition and able to oxidize VOCs at room
temperature.
# 2005 Elsevier B.V. All rights reserved.

Keywords: Cobalt oxide; Ozone decomposition; CO oxidation; VOCs complete oxidation

1. Introduction removed. The main method for neutralization of waste gases

Catalytic processes are used for removing harmful
components from waste gases. It is important to reduce the
temperature of the catalytic reactions with a view to saving
energy, at the same time keeping a high catalytic activity.
Therefore, investigations are aimed at finding effective low-
temperature catalysts or new catalytic processes. Such an
appropriate process is ozone catalytic oxidation (OZCO
method), which uses ozone as an oxidant in catalytic oxidation
reactions. Ozone is known as a strong oxidizing agent for waste
and drinking water treatment, sterilization and deodoration [1–
3]. During the exploitation of copiers and laser printers, ozone
is released into the office air [4]. Due to the fact that ozone itself
is toxic, the residual ozone from these processes must be
* Corresponding author. Tel.: +359 2 979 39 12; fax: +359 2 870 50 24.
E-mail address: petya_konova@abv.bg (P. Konova).
0926-860X/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2005.09.027
containing residual ozone is the heterogeneous catalytic
decomposition. The catalytic ozonation is a promising method
ensuring the purification of waste gases [5–10] and waters [11–
13] from organic contaminants. The main hypothesis is that the
catalytic ozone decomposition yields a highly reactive oxygen
species, which completely oxidize the toxic compounds at
lower temperatures [5,7–10].
The activity of various types of catalysts has been
investigated with respect to both catalytic decomposition of
ozone and catalytic oxidation of CO and VOCs with ozone. In
many cases, alumina-supported metal oxides of Fe, Ag, Co, Ni,
Mn, and Cu have shown high activity for ozone decomposition
at ambient temperature [14,15]. The observed higher ozone
decomposition activity of MnOx/Al2O3 compared to the other
catalysts has been explained by the ability of the manganese
species to reach several different oxidation states [16].
However, the catalytic activity of these metal oxides in
reaction of ozone decomposition decreases with time [17].
110 P. Konova et al. / Applied Catalysis A: General 298 (2006) 109–114
Among transition metals oxides, cobalt oxides, both
unsupported and supported on different oxide support materials
(including alumina), belong to the most important catalysts for
complete oxidation [18]. Their high catalytic activity in air
pollution control of CO [19–22], NOx [23–26] and control of
organic pollutants from effluent streams [27,28] has been
reported in a large number of papers. In previous publications
[29,30], we have investigated a cobalt oxide system, prepared by
the oxidation–precipitation method, which contained an over-
stoichiometric amount of oxygen and Co ions in a high oxidation
state. It has been found that the obtained system, denoted as
CoOx, has a high catalytic activity towards some oxidation
reactions taking place in aqueous and in a gaseous phase [31–33]
as well as towards ozone decomposition (our unpublished data).
According to the results obtained, the synthesized cobalt oxide
system is prominent and useful in a wide field of applications in
environmental protection. From a practical point of view, it is of
interest to investigate the catalytic properties of this oxide system
upon deposition on support materials.
The aim of the present work is to investigate the catalytic
activity of an alumina-supported cobalt oxide system with
overstoichiometric oxygen (denoted as CoOx/Al2O3) during
heterogeneous catalytic decomposition of ozone. Based on the
high oxygen content in this catalyst, the oxidation of CO and
organic substances, using two different oxidants (ozone and
oxygen) will be studied.
2. Experimental
2.1. Sample preparation
The CoOx/Al2O3 catalyst was prepared by deposition
oxidation–precipitation method in an aqueous solution. The
active phase of the supported catalyst (CoOx) was synthesized
according to the procedure described in detail in Refs. [29,30].
The required amount of Co(NO3)26H2O was dissolved in
150 ml of deionized water. The solution was heated up to 70 8C
and then, the support—g-Al2O3 was added and kept under
continuous stirring for 4 h. After cooling to room temperature,
the impregnated catalyst precursor was separated from the
solution and was added to a mixture of aqueous solutions of
NaOH (4 M) and NaOCl (1 M). The solid was kept digested for
24 h, washed several times (until disappearance of the
chlorides), and then dried in an oven at 110 8C for 12 h.
2.2. Sample characterization
The bulk unsupported CoOx catalyst was characterized by
means of IR, XPS, ESR, XRD, DTA and chemical analyses.
The specific surface areas of the supported catalyst and the g-
Al2O3 support were determined by the BET method using low-
temperature nitrogen adsorption. Pore size distribution was
evaluated from the adsorption curve of the isotherm, using the
procedure developed by Orr and Dalla Valle [34]. The chemical
composition was determined by atomic absorption spectro-
metry with Pye Unicam SP90V. The total overstoichiometric
oxygen or so-called active oxygen content, O* (the active
oxygen content is considered to be the amount of oxygen above
the stoichiometric content in the oxide, corresponding to the
lowest stable valence state [35]) was determined by a direct
iodine method [36]. The infrared spectra in the region of 4000–
400 cm
1 was recorded on a Bruker model IFS 25 Fourier
transform interferometer (resolution < 2 cm
1) using KBr
discs as matrices. The XPS studies were performed in a VG
Escalab II electron spectrometer using Al Ka radiation with
energy of 1486.6 eV. The residual gas pressure in the analysis
chamber was 10
7 Pa. Binding energy values (BE  0.2 eV)
were referenced to the C 1s line at 285.0 eV as well as to the Al
2p line of the alumina support at 74.3 eV. The integrated peak
area of the Co 2p line was normalized to that of the relevant Al
2p line corrected by the corresponding photoionization cross
sections. X-ray powder diffraction (XRD) measurements were
carried out using a DRON-3 apparatus with a Co Ka radiation
with scintillation calculation. Magnetic measurements were
carried out in air with a magnetic apparatus constructed
according to the Faraday method. The magnetic susceptibility
(x) was measured in the temperature range of 290–490 K with a
magnetic field intensity varying from 2  103 to 10  103 Oe.
2.3. Catalyst activity measurements
The catalytic activity of the samples was investigated in an
isothermal plug flow reactor. A catalyst particle size of 0.3–
0.5 mm was chosen taking into account the reactor diameter
(6.0 mm) and the volume rate (mean value 22 000 h
1, calculated
with respect to the total catalyst bed volume) in order to reduce the
effect of external diffusion. The preliminary treatment of the
catalysts included heating at 120 8C for 1 h in air flow. The rate of
the gas flow was 4.4 l/h, the catalyst volume—0.2 cm3, and the
mass of the catalyst under consideration—0.15 g.
Ozone was synthesized in an oxygen flow (99.7% dried over
silica gel) using an ozone generator with a silent discharge of 4–
6 kV between the electrodes. The inlet concentration of ozone
for the reaction of ozone decomposition varied between 22.0
and 24.0 g/m3. The ozone concentration was analyzed with an
Ozomat GM (Germany) ozone analyzer with an accuracy of
0.1 g/m3. The reaction temperature varied between
45
and 25 8C and was maintained with an accuracy of 0.2 8C.
The catalytic oxidation of CO and iso-propanol were carried
out within the range of 25–110 8C and 25–250 8C, respectively.
The amounts of CO and iso-propanol were dosed by an Ismatex
MS2/6 (Switzerland) pump. The oxidizing agent was oxygen
from synthetic air (gas mixture of 80% nitrogen and 20%
oxygen) or ozone produced in oxygen. The inlet concentration
of carbon monoxide was 0.18 vol.% and that of iso-propanol—
0.095 vol.%. The inlet concentrations of ozone for CO and iso-
propanol oxidation were chosen in stoichiometric ratios,
considering that only one oxygen atom from the ozone
molecule is consumed. The rate of complete oxidation was
evaluated by measuring the amount of CO2 formed during the
reaction with a Maihak (NDIR) gas analyzer. The CO and CO2
concentrations were determined with an accuracy of 2 ppm.
The same analyzer measured the oxygen with an accuracy of
100 ppm.
 P. Konova et al. / Applied Catalysis A: General 298 (2006) 109–114 111

3. Results and discussion

The results obtained for the bulk unsupported CoOx catalyst

show that the adopted method of synthesis yields an amorphous
and highly disperse CoOx phase with non-stationary surface
properties, high content of active oxygen, high oxidation degree
(Co3+, Co4+) and octahedral coordination of the cobalt ions. Due
to all these attractive properties, one can suppose a high catalytic
activity for the supported CoOx/Al2O3 catalyst in reactions of
complete oxidation in the low-temperature region. The supported
CoOx/Al2O3 catalyst has a larger surface area than the
unsupported one and is thus expected to show a higher activity.
To check this assumption, the supported CoOx/Al2O3 catalyst
was characterized by means of different advanced techniques.
The results of chemical analysis and textural characteriza-
tion of the supported catalyst are listed in Table 1. The surface
area measurements data show that the CoOx deposition induces Fig. 1. Pore size distribution of CoOx/Al2O3 catalyst.
a slight decrease of the BET surface area of the initial support,
indicating a partial blockage of the alumina pores with a pure support (a) and, in accordance with the literature data [37],
smaller diameter. Fig. 1 presents the pore size distribution of the is assigned to the stretching vibration of the surface cobalt–
CoOx/Al2O3 catalyst, which is typical of mesoporous materials. oxygen bond. The shoulder at 725 cm
1 is ascribed to the
The results of the chemical analysis reveal that the deposition of contribution of the alumina support (a), showing in its IR
the active phase on g-Al2O3 yields cobalt oxide with lower spectrum a broad and diffuse absorption hump spreading from
active oxygen content as compared to the unsupported CoOx 400 to 900 cm
1. The band at 580 cm
1 also accounts for the
system, suggesting some surface reduction of the active phase presence of active oxygen in the samples, thereupon its
upon deposition on the support. The XRD analysis shows that intensity is proportional to the content of O*. The spectral range
the supported catalyst is amorphous. of its appearance is evidence of the covalent character of the
The IR spectra of the g-Al2O3 support (a), the bulk phase oxygen bonding to the metal ions of the surface, which in its
CoOx (b), the freshly prepared CoOx/Al2O3 catalyst (c) and the turn justifies the assumption about a high activity of the
used catalyst after ozone decomposition (d) are presented in synthesized sample in reactions of complete oxidation. The
Fig. 2. The spectra of (c) and (d) are identical, showing that the indicated band is similar to the OB3 vibrations in the spinel
catalyst structure is not altered during the catalytic reaction. A lattice of Co3O4 (where B denotes the Co(III) ions in an
broad band in the region 3600–2600 cm
1 (O–H stretching) with octahedral coordination), suggesting that the cobalt in CoOx/
a maximum for spectrum (b) centred about 3410 cm
1, together Al2O3 is situated in an octahedral oxygen environment.
with a band at 1640 cm
1 (H–O–H bending) are observed in all Some differences concerning the position and the int-
spectra. The band centered at 3410 cm
1 can be attributed to the ensity of the band ascribed to the stretching vibration of the
lattice vibration modes of hydrogen bonded hydroxyl groups,
whereas the band at 1640 cm
1 corresponds to adsorbed
molecular water. These results indicate that all the samples
are hydrated/hydroxylated. It is evident from the IR spectra that
the hydrate in the samples is not bonded as oxyhydroxide
(CoOOH should display a sharp peak around 3600 cm
1).
An intense absorption band at 580 cm
1 and a shoulder at
725 cm
1 are observed in the spectra of the samples (c) and (d).
The band at 580 cm
1 is not present in the IR spectrum of the

Table 1
The results of chemical analysis, magnetic measurement and textural char-
acterization of the supported catalyst
BET surface area of the support (m2/g) 325
BET surface area of the catalyst (m2/g) 207
Total pore volume of the catalyst (cm3/g) 0.36
Co loading in the catalyst (%, w/w) 5.16
Co surface density of the catalyst (atoms/nm2) 2.55
O*/Co atomic ratio of the bulk CoOx 1.78
O*/Co atomic ratio of the catalyst 1.51
Magnetic susceptibility of the catalyst x (106)
0.12 Fig. 2. Infrared spectra of: (a) g-Al2O3 support, (b) bulk phase CoOx, (c) freshly
prepared CoOx/Al2O3 catalyst and (d) worked catalyst after ozone decomposition.
112 P. Konova et al. / Applied Catalysis A: General 298 (2006) 109–114
cobalt–oxygen bond are readily observable comparing the IR
spectra of CoOx/Al2O3 samples (c) and (d) with that of the
unsupported CoOx system (b). The deposition of the active
phase on the alumina support causes a shift of this band
towards the lower frequencies (by 10 cm
1), together with a
slight decrease of its integral intensity. These facts indicate
that the ionic feature of the bond is growing stronger, thus
affecting the mobility of the surface oxygen, which has an
impact on the activity of oxide systems in oxidation pro-
cesses. The lower intensity of the band at 580 cm
1 in CoOx/
Al2O3 samples (c) and (d) as compared to the unsupported
CoOx (590 cm
1 (b)) corresponds to the lower active oxygen
content, confirming the results of the chemical analysis.
The Co 2p XPS spectra of fresh and ozonated CoOx/Al2O3
catalysts are shown in Fig. 3, curves (a) and (b), respectively.
The spectra of both samples show similar features, indicating
that upon catalytic reaction there is no change in the electronic
density of the cobalt atoms. A little difference was found for the
values of the Co/Al intensity ratio for fresh and worked
samples—0.171 and 0.154, respectively. The slight decrease of
this ratio for the worked sample might be attributed to some
agglomeration of the loaded cobalt phase. The XPS spectra
contain the spin-orbital components Co 2p3/2 at 780.3 eV and
Co 2p1/2 at 795.3 eV. It is difficult to determine the valence state
of the cobalt ions on the surface of the studied samples by
means of the binding energy alone because its value is quite
close to that for the Co(III) oxides CoOOH (780.0 eV) and
Co2O3 (799.9 eV), as well as to the Co 2p3/2 binding energy for
CoO (780.0 eV) [38]. In this regard, the information about
satellites and the spin-orbit splitting of the 2p level (DE) could
be very useful. Both spectra do not exhibit a prominent satellite
structure, which is characteristic of the Co2+ oxidation state.
However, in the studies of supported catalysts, the information
about the satellite is sometimes ambiguous, especially when the
concentration of the catalytic component is low. The
Fig. 3. XPS spectra of: (a) fresh and (b) used in reaction of ozone decom-
position samples of the CoOx/Al2O3 catalyst.
determined 2p3/2–2p1/2 spin orbit splitting of 15.0 eV in both
spectra is typical of diamagnetic cobalt (Co3+) compounds. The
measured binding energy values, spin orbit splitting and
ambiguous satellites imply that the studied samples contain
cobalt in a Co3+ oxidation state only, present as Co2O3. This
assumption is confirmed by the results of the magnetic
measurements (Table 1), according to which the CoOx/Al2O3
contains Co(III) (octahedral) ions having t62 g configuration and
being diamagnetic (x =
0.12  10
6). The Co 2p XPS
spectrum of CoOx/Al2O3 shows some dissimilarity to these
already observed for the unsupported paramagnetic Co-oxide
system (having a Co 2p3/2 peak at 780.9 eV with its shake-up
satellite at about 6.7 eV to higher binding energies, and DE
value of 14.8 eV).
On the basis of the obtained results from chemical analysis,
magnetic measurements, XPS and IR techniques it may be
concluded that the deposition of CoOx on alumina causes a
slight decrease in the amount of the active surface oxygen and
leads to a lower oxidation state of cobalt ions—mainly Co3+.
According to our preliminary tests on ozone decomposition
and oxidation of organic compounds with oxygen, the bulk-
unsupported system shows a very high activity at room
temperature. The activity of the CoOx phase supported on
alumina remains unchanged, the supported catalysts having a
higher surface area and better mechanic properties than the bulk
phase. Fig. 4 depicts the temperature dependence of the catalyst
activity towards decomposition of ozone, CO and iso-propanol
oxidation using two different oxidation agents—O3 and O2. It is
evident that the activity of the CoOx/Al2O3 catalyst towards
ozone decomposition is very high, even at low temperature
(
45 8C) and reaches almost 100% at room temperature. On
the other hand, the activity shows a tendency to be stabilized
and even to increase with time at a constant temperature. Fig. 5
presents the time–conversion dependence measured at 700 min
at three different temperatures: 25, 0 and
45 8C. One can see
Fig. 4. Conversion–temperature dependences for: (a) ozone decomposition, (b)
CO oxidation with O3, (c) i-propanol complete oxidation with O3, (d) CO
oxidation with O2 and (e) i-propanol complete oxidation with O2 over CoOx/
Al2O3 catalyst.
 P. Konova et al. / Applied Catalysis A: General 298 (2006) 109–114
that at each temperature the activity is lower at the beginning
followed by increase and stabilization with time. This
behaviour reveals that active species develop on the surface
of the CoOx/Al2O3 catalyst during the catalytic ozone
decomposition, which leads to enhancement of the initial
activity.
According to earlier investigations [39], the mechanism of
catalytic ozone decomposition can be presented as follows:
O3 þ Z ! O
Zþ þ O2 (1)
?3 þ ?
Zþ ! 2O2 þ Z (2)
2O
Zþ þ ?2 ! 2Z þ 2O2 (3)
2O
Zþ ! ?2 þ 2Z (4)
The proposed mechanism reveals that for a high activity
towards ozone decomposition the catalyst should have active
sites (Z) capable to change their oxidation state. Such types of
catalysts are the transient metal oxides where several oxidation
states are possible.
It is known that the Co4+ ions are very strong oxidants in
reactions of complete oxidation of VOCs and CO. But this
highest oxidation state of the cobalt is unstable and can be
achieved by using strong oxidants during the synthesis or during
the reaction. It is most likely that in presence of ozone some part
of the Co3+ ions can turn into Co4+ ions. Then, the mechanism of
ozone decomposition can be described as follows:
?3 þ ½Co3þ  ! ?2 þ O
½Co4þ  (a)
?3 þ O
½Co4þ  ! 2O2 þ ½Co3þ  (b)
The first step (a) includes formation of an oxidized state of
the active site—in our case O
[Co4+], which can be named
‘‘catalytically active complex’’ (CAC). During the second step
(b), regeneration of the initial state of E!E occurs. The
prevalent amount of ozone decomposes through reaction (a),
because the formation of O
[Co4+] sites depends on the rate of
Fig. 5. Conversion–time dependence of the CoOx/Al2O3 catalyst in ozone
decomposition at different temperatures.
113
reaction (a). In the presence of oxygen, another two parallel
reactions can proceed:
2O
½Co4þ  þ ?2 ! 2½Co3þ  þ 2O2 (c)
2O
½Co4þ  ! ?2 þ 2½Co3þ  (d)
Obviously, in the beginning of the process of ozone
decomposition, the concentration of O
[Co4+] sites is negligibly
low and the decomposition of ozone is due to reaction (a). As the
reaction proceeds, the concentration of O
[Co4+] sites increases
and the participation of reaction (b) becomes appreciable, which
results in a higher degree of conversion.
On the basis of the obtained results on the reaction of ozone
decomposition, ozone is used as an oxidizing agent and
compared with oxygen as an oxidant for the oxidation of CO
and iso-propanol. The conversion–temperature dependences
(Fig. 4) show that when ozone is used, a better conversion can
be achieved and at the same time the oxidation temperature of
the different substances significantly decreases. It is evident
that the temperature for 50% conversion of iso-propanol in
presence of ozone is lower with 185 8C as compared to the case
of oxidation in air. In the case of CO oxidation at 80 8C in
presence of ozone the conversion is 60%, whereas in air at the
same temperature only 7% can be achieved. Comparison of the
catalytic activity at temperatures higher than 80–85 8C for
oxidation reactions with ozone is not correct due to the
increased effect of homogeneous ozone decomposition [40].
One can see that with ozone, in comparison with oxygen as
an oxidant, the degree of CO conversion is lower than the
oxidation of i-propanol at a given temperature. Most probably,
this is due to the fact that the oxidation with ozone has a
mechanism, different from that of the oxidation with oxygen.
The active sites for oxidation with ozone are O
[Co4+], which
are formed during the decomposition of ozone. This type of
active sites is not present on the catalyst surface in the reaction
of oxidation with oxygen.
Therefore, the oxidation with ozone not only leads to higher
catalytic activity at lower temperatures, but also to a change in
the selectivity of the reaction. The high catalytic activity using
ozone as oxidant permits the reaction to be carried out in a low-
temperature region (below 80 8C), which is extremely
important for the catalytic oxidation of toxic volatile organic
compounds, because it permits saving energy. A very important
feature of the CoOx/Al2O3 catalyst is that it possesses both
activity toward ozone decomposition and oxidation with ozone,
which leads to removal of the residual ozone from waste gases.
Hence, the CoOx/Al2O3 catalyst is very suitable for application
in environmental catalysis.
4. Conclusion
 According to the results obtained from chemical analysis,
magnetic measurements, XPS and IR techniques, the
deposition of a CoOx system on alumina causes a slight
decrease in the amount of the active surface oxygen and leads
to a lower oxidation state of the cobalt ions (mainly Co3+)
which have an octahedral configuration and are diamagnetic.
114 P. Konova et al. / Applied Catalysis A: General 298 (2006) 109–114
 The CoOx/Al2O3 catalyst shows a very high activity at room
temperature in both the reactions of ozone decomposition and
oxidation with ozone, the catalyst remaining active with the
time. The catalytic activity with respect to the complete
oxidation of iso-propanol and the oxidation of carbon
monoxide, in presence of ozone is higher than in presence of
oxygen. A significant increase in the catalytic activity and
decrease in the reaction temperature is observed when ozone
is used as an oxidant.
 Two main reasons for the high catalytic activity are found: (i)
the high content of active and mobile oxygen obtained during
the synthesis on the catalyst surface; (ii) the catalytic active
complexes of O
[Co4+], which are formed during the
reaction of ozone decomposition and are able to oxidize
VOCs at room temperatures.
References
[1] S.T. Oyama, Catal. Rev. 42 (3) (2000) 279.
[2] R. Gracia, S. Cortes, J. Sarasa, P. Ormad, J.L. Ovelleiro, Ozone: Sci. Eng.
22 (2) (2000) 185.
[3] G. Centi, S. Perathoner, Catal. Today 53 (1) (1999) 11.
[4] R. Heck, R. Farrauto, H. Lee, Catal. Today 13 (1992) 43.
[5] S. Imamura, M. Ikebata, T. Ito, T. Ogita, Ind. Eng. Chem. Res. 30 (1)
(1991) 217.
[6] Z. Hao, D. Cheng, Y. Guo, Y. Liang, Appl. Catal. B 33 (3) (2001) 217.
[7] D. Mehandjiev, A. Naydenov, G. Ivanov, Appl. Catal. A 206 (2001) 13.
[8] D. Mehandjiev, K. Cheshkova, A. Naydenov, V. Georgesku, React. Kinet.
Catal. Lett. 76 (2) (2002) 287.
[9] A. Naydenov, D. Mehandjiev, Appl. Catal. 97 (1993) 17.
[10] A. Naydenov, R. Stoyanova, D. Mehandjiev, J. Mol. Catal. 98 (1995) 9.
[11] B. Legube, N. Karpel Vel Leitner, Catal. Today 53 (1) (1999) 61.
[12] C. Cooper, R. Burch, Wat. Res. 33 (18) (1999) 3695.
[13] N.K.V. Leitner, B. Delouane, B. Legube, F. Luck, Ozone: Sci. Eng. 21 (3)
(1999) 261.
[14] S.T. Oyama, Catal. Rev. Sci. Eng. 42 (2000) 279.
[15] H. Einaga, S. Futamura, React. Kinet. Catal. Lett. 81 (2004) 121.
[16] R. Radhakrishnan, Y. Ohminami, K. Asakura, J. Phys. Chem. 105 (2001)
9067.
[17] B. Dhandapani, S.T. Oyama, J. Appl. Catal. B: Environ. 11 (1997)
129.
[18] D.G. Castner, Ph.R. Watson, I.Y. Chan, J. Phys. Chem. 93 (1989)
3188.
[19] Y.I. Mergler, A. van Aalast, J. van Delfi, B.E. Nieuwenhuys, J. Catal. 161
(1996) 310.
[20] K. Omata, T. Takada, S. Kasahara, M. Yamada, Appl. Catal. A 146 (1996)
245.
[21] Y.I. Mergler, J. Hoebink, B.E. Nieuwenhuy, J. Catal. 167 (1997) 305.
[22] J. Jansson, J. Catal. 194 (2000) 55.
[23] L.F. Loitta, G. Pantaleo, A. Macaluso, G. Di Carlo, G. Geganello, Appl.
Catal. A 245 (2003) 167.
[24] D. Pietrogiacomi, S. Tuti, M.C. Campa, V. Indovina, Appl. Catal. B 28
(2000) 43.
[25] L. Chen, T. Horiuchi, T. Mori, Catal. Lett. 72 (2001) 71.
[26] T. Horiuchi, T. Fujiwara, L. Chen, K. Suzuki, T. Mori, Catal. Lett. 78
(2002) 319.
[27] E. Garbowski, M. Guenin, M.C. Marion, M. Primet, Appl. Catal. B 64
(1990) 209.
[28] A.K. Sinha, V. Shankar, J. Chem. Eng. Biochem. Eng. 52 (1993) 115.
[29] St. Christoskova, M. Stojanova, M. Georgieva, D. Mehandjiev, Mater.
Chem. Phys. 60 (1999) 39.
[30] St. Christoskova, M. Stojanova, M. Georgieva, D. Mehandjiev, Thermo-
chim. Acta 292 (1997) 77.
[31] St.G. Christoskova, M. Stoyanova, M. Georgieva, Proceeding of the Ninth
Int. Symp. Heterog. Catal., Varna, Bulgaria, 2000), p. 779.
[32] St. Christoskova, M. Stojanova, M. Georgieva, O. Argirov, React. Kinet.
Catal. Lett. 66 (1) (1999) 55.
[33] St.G. Christoskova, M. Stoyanova, G. Ivanov, D. Mehandjiev, Bulg.
Chem. Com. 30 (1–4) (1998) 470.
[34] C. Orr, J.M. Dalla Valle, Fine Particle Measurements, Macmillan, New
York, 1959, p. 271.
[35] D. Pantony, A. Siddiqi, Talanta 9 (1962) 811.
[36] K. Nakagava, R. Konaca, T. Nakata, J. Org. Chem. 7 (1962) 1597.
[37] A. Bielanski, J. Haber, Oxygen in Catalysis, Marcel Dekker Inc., New
York, 1991.
[38] N.S. Mclntyre, M.G. Cook, Anal. Chem. 47 (1975) 2208.
[39] P. Konova, A. Naydenov, P. Dimitrova, D. Mehandjiev, Chem. Tech. 31 (1)
(2004) 10.
[40] S.D. Razumovskii, S.K. Rakovski, D.M. Shopov, G.E. Zaikov, Ozone and
its Reactions with Organic Compounds, Academic Press, Sofia, Bulgaria,
1983.