Characterization and catalytic activity for the NO decomposition and reduction by CO of nanosized Co3O4

Article abstract of DOI:10.1016/j.jallcom.2004.09.013

Nanosized Co₃O₄ prepared by a precipitation method was characterized by TEM, XRD, BET and TPD techniques, and studied for NO decomposition and reduction by CO. It is found that Co₃O₄ thus obtained has a specific surface area of 23.4 m²/g and an average particle size of 26 nm. Catalytic tests showed that full NO conversion to N₂ was obtained above 300 °C. A redox mechanism between Co ³⁺ and Co²⁺ ions based on NO decomposition is proposed.

Full text of DOI:10.1016/j.jallcom.2004.09.013

Journal of Alloys and Compounds 392 (2005) 317–321
Characterization and catalytic activity for the NO decomposition and
reduction by CO of nanosized Co3O4
Zhaoliang Zhanga,∗, Haoran Gengb, Lisheng Zhenga, Bin Dua
a
College of Chemistry and Chemical Engineering, Jinan University, Ji Wei Road 106, Jinan 250022, PR China
College of Materials Science and Engineering, Jinan University, Ji Wei Road 106, Jinan 250022, PR China
b
Received 30 July 2004; received in revised form 14 September 2004; accepted 15 September 2004
Available online 30 October 2004
Abstract
Nanosized Co3O4 prepared by a precipitation method was characterized by TEM, XRD, BET and TPD techniques, and studied for NO
decomposition and reduction by CO. It is found that Co3O4 thus obtained has a specific surface area of 23.4 m2/g and an average particle size
of 26 nm. Catalytic tests showed that full NO conversion to N2 was obtained above 300 ◦C. A redox mechanism between Co3+ and Co2+ ions
based on NO decomposition is proposed.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Cobalt oxide; Nanosized; NO; CO; Catalytic reduction
1. Introduction
for the NO + CO reaction, that is, at 300 ◦C, 100% NO con-
version is reached, but when the temperature increased the
conversion fell to 70% before increasing again to reach 100%
above 535 ◦C. Unfortunately, we did not find this. Precipita-
tion is one of the most widely employed preparation methods
to obtain nanosized materials, which is very easy and cheap
as regards industrialization and commercialization. In this
paper, the nanosized Co3O4 was prepared by precipitation
and characterized by TEM, XRD, BET and TPD techniques.
Its catalytic activity in the reduction of NO by CO is checked
and discussed.
Cobalt oxides have been extensively studied for a long
time in the field of air pollution control. Co3O4 [1] and
cobalt oxides doped with Cu [2], Fe [3], or supported on
␥-Al2O3 [4], are well known to be active catalysts for CO
oxidation at low temperatures. CoMgO solid solution [5] and
LaCoO3 composite oxide [6] are used for methane oxidation
and VOCs combustion, respectively. Besides, bulk Co3O4 is
one of the most active single component metal oxides for NO
decomposition, and its activity can also be enhanced by addi-
tion of Ag and Na [7]. Accordingly, bulk and supported cobalt
oxides have revealed high activity for the reduction of NO by
CO [8] and hydrocarbons [9]. The reduction of NO by CO
is an important reaction for exhaust gas depollution because
the two pollutants are simultaneously eliminated in one step.
Panayotov et al. [2] reported the interaction between NO and
CO over Co3O4, which was prepared from the decomposi-
tion of cobalt nitrate, with a surface area as low as 9 m2/g.
Simonot et al. [8] found a unique phenomenon on Co3O4
2. Experimental
Co3O4 were prepared by a precipitation method. The sto-
ichiometric solutions of Co(NO)2·6H2O and ammonia were
simultaneously dropped into de-ionized water under vigor-
ous agitation, and then the solution was aged for 30 min. The
resultant precipitates were dried at 120 ◦C overnight and cal-
cined at 500 ◦C for 5 h in air.
∗
A U-shaped quartz reactor (i.d. = 6 mm) with a porous
quartz frit supporting the catalyst was used for reaction tests
Corresponding author.
E-mail address: zhangzhaoliang@sdu.edu.cn (Z. Zhang).
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2004.09.013

318
Z. Zhang et al. / Journal of Alloys and Compounds 392 (2005) 317–321
under atmospheric conditions. The model flue gases em-
ployed are 1025 ppm NO and 2085 ppm CO in Ar. The total
flow rate was 40 cm3/min, and 300 mg of catalyst was placed
in the reactor leading to a space velocity of 8000 ml/(h g). The
products in the effluent stream were analyzed by an online
quadrupole mass spectrometry (LZL-204, Beijing Analyti-
cal Instrument Plant). The NO conversion was calculated on
the basis of the differences between the inlet and outlet NO
intensities. The data for steady-state activity of the catalyst
were collected after 2 h testing.
Conversion of NO (%):
[NO intensity]in − [NO intensity]out
XNO
=
× 100
[NO intensity]in
The morphology of the nanosized particles was examined
with a JEM-2010 (JEOL) transmission electron microscopy
at an accelerating voltage of 200 kV.
X-ray powder diffraction (XRD) patterns were recorded
on a Rigaku D/max 2000 diffractmeter employing Cu K␣
radiation. The BET specific surface area and pore structure
were measured in a Micromeritics ASAP-2010 instrument.
The temperature-programmed desorption of NO was car-
ried out on the above equipment. A sample of 100 mg was
first heated in 20 ml/min Ar flow to 500 ◦C with a 15 ◦C/min
ramp, followedbycoolingtoambienttemperatureinthesame
flow. After a short purging, the flow was switched to a mix-
ture of 2050 ppm NO in Ar at a flow of 20 ml/min and main-
tained for 60 min, then the flow was switched back to Ar at
20 ml/min, and the sample was purged for 1 h. A NO-TPD
profile was registered in the same flow while the temperature
was increased to 500 ◦C at 10 ◦C/min.
The transient response technique is used to elucidate NO
decomposition and reaction mechanism. First, Ar flowed
through the catalyst at the desired temperature, and then the
concentration step change from Ar to 2050 ppm NO in Ar was
enforced by switching a four-way valve. The mass-to-charge
(m/e) ratios were monitored by MS as follows: NO (30), O2
(32), CO and N2 (28), CO2 and N2O (44), NO2 (46).
3. Results and discussion
Fig. 1. (a) and (b) TEM images of Co3O4 nanoparticles.
3.1. TEM characterization
The transmission electron microscope images of Co3O4
prepared by precipitation and calcined at 500 ◦C are shown in
Fig. 1. It can be seen that there exists a certain extent of coag-
ulation. The particle size was between 20 and 40 nm (Fig. 1a).
High-magnification image showed clear lattice strings of
crystalline phases and pseudo-hexagonal morphology of the
nanoparticles (Fig. 1b).
ambient temperature to 500 ◦C, as shown in Fig. 2. Dur-
ing the whole range, nearly complete NO conversion to N2
with negligible amount of N2O formation was achieved.
The NO conversion showed a S-shape curve with tem-
perature change. The light-off temperature, at which 50%
NO conversion was reached, is about 110 ◦C. Almost com-
plete NO conversion could be achieved at 300 ◦C, and no
activity decrease was detected thereafter. The CO/NO ra-
tio used was twice the stoichiometric ratio in this reaction
[10].
3.2. Catalytic activity
The catalytic activity of Co3O4 for NO reduction by CO
was measured as a function of reaction temperatures from

Z. Zhang et al. / Journal of Alloys and Compounds 392 (2005) 317–321
319
Table 1
Characterization of calcined Co3O4 at 500 ◦C for 5 h
Characteristics
Values
23.4
0.1338
BET surface area (m2/g)
Pore volume (cm3/g)
˚
Average pore diameter (A)
Crystallite size (nm)
228.5
26
reduced to CoO, an oxygen deficient phase, even though the
ratio of the reacting molecules of NO and CO equals unit.
This observation is evidence that the reaction occurred not
only via the surface but also via an intrafacial process.
3.4. Transient studies
NO decomposition is checked using the transient response
method, which is based on recording the response of a given
parameter of the reaction system to a sharp change in the
steady state. It is traced until a new steady state is attained.
Fig. 4 shows the transient responses obtained at 350 ◦C on
Co3O4 after switching from Ar to 2050 ppm NO in Ar. The
instantaneousN2 responsecurveconcomitantwiththeforma-
tion of a small fraction of N2O indicated the catalyst was ac-
tive in the NO decomposition in the absence of CO in the gas
phase, and that NO dissociated upon adsorption. However,
as no oxygen desorbed from the catalyst, the catalyst surface
would become oxidized. The catalytic activity of NO decom-
position into N2 gradually decreased to zero after 40 min. Al-
ternatively, about 9% NO conversion to N2O was observed.
Fig. 2. NO conversions in the NO + CO reactions.
3.3. XRD characterization
Fig. 3a shows the XRD pattern of the fresh catalyst. The
peak positions agree well with the reflections of bulk spinel
Co3O4. According to the Scherrer equation D = kλ/(β cos θ),
where k is a constant, λ the X-ray wavelength, β the breadth
at half-maximum intensity and θ is the Bragg angle in the
diffraction pattern, the crystallite size D is equal to 26 nm in
diameter, which is in a reasonable agreement with the TEM
results. The surface characteristics of the resultant Co3O4,
namely specific surface area (SBET), total pore volume and
average pore radius were determined from nitrogen adsorp-
tion isotherms measured at −196 ◦C. The results obtained
are given in Table 1, which shows a BET surface area of
23.4 m2/g, a pore volume of 0.1338 cm3/g and an average
3.5. NO-TPD
NO-TPD was performed on Co3O4 for three times,
showing that a large amount of NO can adsorb and desorb
reversibly. Unfortunately, reproducible spectra were never
obtained (not shown here). Moreover, an increasing NO
desorption was recorded after the former cycle. From
transient studies, we know that the interaction between NO
and Co3O4 is the process of Co3O4 oxidization by NO,
which will result in a change in oxidation state of some
low-valent Co ions or a surface reconstruction [12] after
NO-TPD, thus different desorption spectra with respective
desorption temperatures and amounts were observed. It is
confirmed that there was more NO sorption capacity on the
oxidized catalyst surface [2], which can explain more NO
desorption for the later NO-TPD cycle than the former.
˚
pore diameter of 228.5 A. These data meant that the pores
were the voids among the dense coagulation of original parti-
cles, as shown in Fig. 1a. For comparison, the XRD pattern of
the used catalyst after NO + CO reactions is shown in Fig. 3b.
In accordance with the previous work [11], Co3O4 was partly
3.6. Reaction mechanism
It was previously observed that preoxidized cobalt oxide
catalysts show high low-temperature activity for CO oxida-
tion. While the activity over prereduced cobalt oxide is much
lower. Therefore, CO was suggested to be adsorbed on an
oxidized cobalt site, probably Co3+ [4]. The adsorbed CO
reacts with oxygen linked to the active Co3+; CO2 is formed
and desorbed quickly. The result is a partially reduced site,
Fig. 3. XRD patterns of the catalyst before and after NO + CO reactions.

320
Z. Zhang et al. / Journal of Alloys and Compounds 392 (2005) 317–321
Fig. 4. Transient response curves obtained over Co3O4 at 350 ◦C after switching from Ar to 2050 ppm NO in Ar.
which may consist of two Co2+ ions or may be regarded as an
oxygen vacancy. On the basis of the mechanism of NO de-
composition on metal oxides, it is reasonable to deduce that
NO adsorbed on Co2+ or oxygen vacancy sites. Thus, the fol-
lowing elementary steps are proposed for NO decomposition
and reduction by CO on Co3O4, in which stands for oxygen
vacancies, “ad” and “L” are the abbreviations of adsorption
and lattice, respectively:
Overall, the reaction between NO and CO can be written
as
2NO + 2CO → N2 + 2CO2
(7)
4. Conclusions
Nanosized Co3O4 was prepared by precipitation of a so-
lution of cobalt nitrate using ammonia, which is easy and
cheap as regards industrialization and commercialization. A
crystallite size of 26 nm and BET surface area of 23.4 m2/g
were reached.
Using CO as a reducing agent, Co3O4 showed complete
NO conversion to N2 above 300 ◦C. The reaction took place
simultaneously with a reduction of Co3O4 into CoO. A re-
dox mechanism between Co3+ and Co2+ ions based on NO
decomposition was proposed.
NO + Co2+ → Co3+NO−
(1)
(2)
(3)
(4)
(5)
(6)
Co3+NO− → Co3+Oad− + Nad
Nad + Nad → N2 ↑
Nad + Co3+NO− → N2O ↑ +Co2+
Co3+Oad− + CO → Co2+ + CO2
Co3+Oad− + CO → Co2+ + CO2
Acknowledgment
In the absence of CO, when the oxygen vacancies were not
enough or depleted by the reaction, there was not sufficient
This research was supported, in part, by PhD Fund from
Jinan University (B0403).
Nad to ensure the occurrence of reaction (3). On the contrary,
reaction (4) will prevail, as shown by the NO decomposi-
tion (Fig. 4). However, because Oad would not be desorbed,
Co3O4 shows no sustainable activity for NO decomposition
to N2 and/or N2O. When CO was added, that is, the reduc-
tion of NO by CO, reaction (5) produced enough oxygen
vacancies for NO adsorption and decomposition, and thus a
reaction cycle was constructed. Simultaneously, reaction (6)
would result in the production of oxygen vacancies available
for reaction (1) and in a partial catalyst reduction to CoO.
References
[1] L. Simonot, F. Garin, G. Maire, Appl. Catal. B 11 (1997) 167.
[2] D. Panayotov, M. Khristova, D. Mehandjiev, Appl. Catal. 34 (1987)
49.
[3] V.K. Yatsimirskii, E.V. Ishchenko, N.A. Boldyreva, N.N. Khomenko,
G.G. Tsapyuk, Theor. Exp. Chem. 37 (2001) 315.
[4] J. Jansson, J. Catal. 194 (2000) 55.

Z. Zhang et al. / Journal of Alloys and Compounds 392 (2005) 317–321
321
[5] S.F. Jia, T.C. Xiao, H.T. Wang, E. Flahaut, K.S. Coleman, M.L.H.
Green, Catal. Lett. 75 (2001) 65.
[9] T. Horiuchi, T. Fujiwara, L. Chen, K. Suzuki, T. Mori, Catal. Lett.
78 (2002) 319.
[6] R. Spinicci, M. Faticanti, P. Marini, S. De Rossi, P. Porta, J. Mol.
Catal. A 197 (2003) 147.
[7] P.W. Park, J.K. Kil, H.H. Kung, M.C. Kung, Catal. Today 42 (1998)
51.
[10] F. Solymosi, J. Kiss, J. Catal. 54 (1978) 42.
[11] D. Panayotov, M. Khristova, D. Mehandjiev, Appl. Catal. 34 (1987)
65.
¨
[12] J. Jansson, A.E.C. Palmqvist, E. Fridell, M. Skoglundh, L. Osterlund,
[8] L. Simonot, F. Garin, G. Maire, Appl. Catal. B 11 (1997) 181.
P. Thorma¨hlen, V. Langer, J. Catal. 211 (2002) 387.