Cobalt-iridium impregnated zirconium-doped mesoporous silica as catalysts for the selective catalytic reduction of NO with ammonia

Article abstract of DOI:10.1016/j.molcata.2005.12.019

Catalysts based on cobalt-iridium supported on zirconium-doped mesoporous silica, with a Si/Zr molar ratio of 5, a cobalt loading of 16 wt.% and Co/Ir molar ratios of 10, 20, 30 and 60 were prepared by impregnation, characterised and then tested in the selective catalytic reduction of NO with ammonia in an excess of oxygen. These catalysts are active in the targeted catalytic reaction showing high conversions of NO at low reaction temperatures with a low yield of N₂O. The incorporation of iridium increases the dispersion of cobalt, thus improving the catalytic performance. These catalysts maintain their activities when water is added to the feed on stream, but are less active with the addition of 100 ppm of SO₂.

Full text of DOI:10.1016/j.molcata.2005.12.019

Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
Cobalt–iridium impregnated zirconium-doped mesoporous silica as catalysts
for the selective catalytic reduction of NO with ammonia
∗
´
´
´
´
´
Ramon Moreno-Tost, Enrique Rodrıguez Castellon, Antonio Jimenez-Lopez
Departamento de Qu´ımica Inorga´nica, Cristalograf´ıa y Mineralog´ıa (Unidad Asociada al ICP-CSIC),
Facultad de Ciencias, Universidad de Ma´laga, Campus de Teatinos, 29071 Ma´laga, Spain
Available online 24 January 2006
Abstract
Catalysts based on cobalt–iridium supported on zirconium-doped mesoporous silica, with a Si/Zr molar ratio of 5, a cobalt loading of 16 wt.%
and Co/Ir molar ratios of 10, 20, 30 and 60 were prepared by impregnation, characterised and then tested in the selective catalytic reduction of NO
with ammonia in an excess of oxygen. These catalysts are active in the targeted catalytic reaction showing high conversions of NO at low reaction
temperatures with a low yield of N2O. The incorporation of iridium increases the dispersion of cobalt, thus improving the catalytic performance.
These catalysts maintain their activities when water is added to the feed on stream, but are less active with the addition of 100 ppm of SO2.
© 2005 Elsevier B.V. All rights reserved.
Keywords: NO; Selective catalytic reduction; Iridium; Cobalt; SO2; XPS; NO-TPD; Mesoporous silica; Zirconium
1. Introduction
but an active catalyst for industrial applications has not yet been
obtained. The most commonly used catalysts are based on mixed
oxides: V2O5–WO3–TiO2 for the SCR of NO with ammonia.
However, these catalysts suffer from some disadvantages: their
activity is highest at temperatures above 250 ◦C, the toxicity of
vanadium, etc. For these reasons, over recent decades, many
research groups have been studying other systems for supports
and active phases which are effective in the SCR of NO with
ammonia. Research has been carried out on catalysts based on
zeolites [12,13], SiO2 [14], Al2O3 [15,16], etc., and on others
loaded with different metals such as Fe [13,14], Cu [12,17], Cr
[18,19] or Mn [15,20].
Since Li and Armor [21] reported that Co-catalysts could
reduce the NO to nitrogen with methane in the presence of oxy-
gen, the use of cobalt as an active phase has been limited mainly
to the presence of methane as a reductant. Other catalysts are
also active in the catalytic reduction of NO with methane such
as Ga [22], Ir [23] or Pd [24]. However, the Co-catalysts are
not stable in the presence of water vapour or SO2 [25,26]. To
overcome these undesired effects, bimetallic catalysts have been
synthesised and tested in the SCR of NO with hydrocarbons.
Supported bimetallic catalysts are of great interest because one
metal can tune or modify the catalytic properties of another as
a result of both electronic and structural influences. Normally,
the second metal added to the Co-catalyst is a noble metal such
as Pd [27] or Pt [28]. The presence of Pt or Pd appears to pro-
duce an increase in the amount of NO adsorbed, and may also
One of the main environmental problems that concerns the
scientific community today is related to nitrogen oxides (NOx)
emitted into the atmosphere. NOx is one of the most harmful
pollutants involved in acid rain as well as playing a major role
in the formation of trophospheric ozone with volatile organic
compounds, etc. N2O is responsible for global warming and the
destruction of stratospheric ozone. Restrictive legislation has led
to the control of these emissions around the world [1].
NO is produced during combustion processes by the oxida-
tion of atmospheric nitrogen at very high temperatures. Some
NO is also formed by oxidation of nitrogen compounds present
in fuel [2]. Thus, NO formation may be controlled either by
improving the combustion process or by reducing its concen-
tration in the flue-gases. The latter can be achieved with cat-
alytic technology such as the selective catalytic reduction (SCR),
where a reductant such as ammonia [3,4], CO [5,6] or hydro-
carbons [7,8] are necessary to convert NO into N2. Nowadays,
for stationary sources of emission, the only available method
to reduce the NO concentration in waste gases on an industrial
scale is the SCR of NO with ammonia. Recently, much effort has
been devoted to replacing ammonia with hydrocarbons [9–11],
∗
Corresponding author. Tel.: +34 952 131876; fax: +34 952 137431.
E-mail address: ajimenezl@uma.es (A. Jime´nez-Lo´pez).
1381-1169/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2005.12.019

R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
127
be responsible for NO oxidation to NO2 (an important reaction
2.2. Characterisation methods
intermediate for HC-SCR) and water tolerance. Nevertheless,
the Co-catalysts have not been tested in depth in the SCR of NO
withammonia. Thereareonlyafewreportsbasedonco-catalysts
[29–31] giving worse results than those reported when methane
is used as a reductant. Chmielarz et al. [32] found that a titania-
pillared montmorillonite exchanged with cobalt achieved a NO
conversion close to 100% at temperatures higher than 400 ◦C,
whilst displaying high selectivities towards nitrogen. Co-ZSM-
5 catalysts prepared with different methods have been assayed
with iso-butane as reductant showing high yield of nitrogen at
lowtemperatureforcatalystscontainingCo2+oxo-ionsorCo3O4
clusters [33].
Recently, catalysts based on iridium have received much
more attention since they exhibit high activity in the reduc-
tion of NO in excess oxygen [34,35] with different reductant
agents. Ogura et al. [36] reported that Ir-silicate exhibits high
activity for the NO–CO reaction. Wang et al. [37] have reported
that Ir-ZSM-5 monoliths could efficiently reduce NO with CO
under lean conditions. Yoshinari et al. [38] reported that Ir–SiO2
showed excellent activity for NO reduction with hydrogen in the
presence of O2 and SO2. Moreover, several authors have empha-
sised the promotional effect of SO2 present in the feed stream
[38,39].
Powder XRD patterns were obtained with a Siemens D500
diffractometer, equipped with a graphite monochromator and
using Cu-K␣ radiation. X-ray photoelectron spectra were col-
lected using a Physical Electronics PHI 5700 spectrometer
with non–monochromatic Mg K␣ radiation (300 W, 15 kV,
1253.6 eV) for the analysis of the photoelectronic signals of Si
2p, Zr 3d, O 1s, Co 2p and Ir 4f, and equipped with a multi-
channeldetector. TheCo2pphotoelectronicsignalwascollected
by using a non-monochromatic Al K␣ radiation (300 W, 15 kV,
1486.6 eV) as an excitation source to avoid the overlapping of
the O KLL Auger line with the photoelectronic Co 2p signal.
Spectra of powdered samples were recorded with the constant
pass energy values at 29.35 eV, using a 720 ␮m diameter analy-
sis area. During the data processing of the XPS spectra, binding
energy values were referenced to the C 1s peak (248.8 eV) from
the adventitious contamination layer. The PHI ACCESS ESCA-
V6.0 F software package was used for acquisition and data anal-
ysis. A Shirley-type background was subtracted from the sig-
nals. Recorded spectra were always fitted using Gauss–Lorentz
curves in order to determine the binding energy of the differ-
ent element core levels more accurately. The error in BE was
estimated to be ca. 0.1 eV.
Our previous work related to Co-catalysts [40] used in the
SCR of NO with ammonia showed that cobalt could be an excel-
lent active phase with high activities and selectivity towards
nitrogen. In order to improve its catalytic activity in the presence
of water vapour and SO2 in the feed stream, iridium has been
selected as the promoter because the presence of iridium may
improve the catalytic activity of the cobalt catalyst by increas-
ing the dispersion of cobalt oxides and therefore the number of
active sites. Thus, a family of Co/Ir catalysts supported on a zir-
conium doped-mesoporous silica with a Si/Zr molar ratio of 5
was prepared with a view to testing its catalytic activity in the
SCR of NO with ammonia in the presence of water vapour and
SO2.
Textural parameters were obtained from N2 adsorption–des-
orption isotherms (BET method) as determined with a Micro-
meritics ASAP 2020. Temperature-programmed reduction of H2
(H2-TPR) was performed between room temperature and 800 ◦C
using a flow of Ar/H2 (40 cm3 min−1, 10 vol.% of H2) and a
heating rate of 10 ◦C min−1. Water produced in the reduction
was eliminated by passing the gas flow through a cold finger
(−80 ◦C). The consumption of H2 was monitored by an on-line
gas chromatograph (Shimadzu GC-14) provided with a TCD.
Temperature-programmed desorption of NO (NO-TPD) was
performed by adsorbing NO onto the catalysts at room tempera-
ture (150 cm3 min−1 flow rate and 0.05 vol.% NO balanced with
He) for 1 h, and performing desorption between 40 and 550 ◦C,
using a heating rate of 10 ◦C min−1. Before the adsorption of
NO, the catalysts were heated at 550 ◦C under a He flow for 1 h.
During desorption, helium was flushed through and the eluted
gas phase was carefully monitored and quantified using a Balz-
ers GSB 300 02 on-line quadrupole mass spectrometer.
2. Experimental
2.1. Preparation of catalysts
A zirconium-doped mesoporous silica, with a Si/Zr molar
ratio of 5, was prepared by following the method described in
previous papers [41,42]. The material was calcined at 550 ◦C
(1 ◦C min−1 heating rate) for 6h and was denoted as SiZr5. This
material was used as a support with a cobalt loading of 16 wt.%
and Co/Ir atomic ratios of 10, 20, 30 and 60 which were pre-
pared by the incipient wetness method. Both metals were added
simultaneously to the support. After impregnation, the samples
were calcined at 550 ◦C for 4 h (1 ◦C min−1 heating rate). For
comparison, two monometallic catalysts with a cobalt loading
of 16 wt.% and a iridium loading of 2 wt.% were prepared,
following the same method. The monometallic catalysts were
named as Co-16 and Ir-2 and the bimetallic catalysts as CoIr10,
CoIr30, CoIr20 and CoIr60, depending on the Co/Ir atomic
ratios.
2.3. Catalysis
Catalysts were tested in the SCR of NO by using a Pyrex
glass tube microreactor (6.85 mm o.d.) working at atmospheric
pressure in a steady-state flow mode and with a catalytic charge
of 150 mg of pelletised solids, sieved to 0.3–0.4 mm, in all cases
without dilution. Samples were pretreated at 350 ◦C in situ for
2 h under a He flow (30 cm3 min−1). The gas reaction mixture
was composed of 1000 ppm NO, 1000 ppm NH3 and 2.5 vol.%
O2 (balanced with helium). In some tests, 10 vol.% H2O, passing
helium through a saturator with deionized water and 100 ppm
of SO2, was added to the feed stream. The flows were indepen-
dently controlled by channel mass flowmeters (Brooks) and a
total flow rate of 42 cm3 min−1 was used in the feed. The space

128
R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
Table 1
Textural properties CoIr catalysts
Catalyst
Ir (%)
SBET
(m2 g−1
Pore
diameter (A)
Pore volumea
a
(cm3 g−1
)
˚
)
SiZr5
Co-16
Ir-2
CoIr60
CoIr30
CoIr20
CoIr10
–
–
632
540
602
514
509
473
473
33.4
36.0
30.2
32.6
32.3
36.1
35.3
0.605
0.618
0.583
0.541
0.519
0.541
0.523
2.6
0.9
1.8
2.6
5.2
a
Determinated by BJH method.
velocity (GHSV) was 3300 h−1; in these conditions, both exter-
nal and internal diffusional limitations were absent. The reaction
was studied from 100 to 500 ◦C. The analysis of reactants and
products (NO, N2O, N2, NH3, H2O and SO2) was monitored
by using the on-line quadrupole mass spectrometer described
above. In previous experiments, variation in the amount of cata-
lystwithatotalflowratemaintainingthespacevelocityconstant,
produced no modification of conversion values. No influence of
the particle diameter was found either.
Fig. 1. Powder XRD patterns of (a) Ir-2, (b) Co-16, (c) CoIr10, (d) CoIr20, (e)
CoIr30 and (f) CoIr60.
3. Results and discussion
3.1. Characterisation of catalysts
A family of bimetallic catalysts were prepared with an atomic
ratio Co/Ir of 10, 20, 30 and 60, but with the same cobalt loading
of 16 wt.% supported on a zirconium doped mesoporous silica.
Moreover, two monometallic catalysts with a cobalt loading of
16 wt.% and iridium loading of 2 wt.% were synthesised for
comparison. Table 1 summarises the textural properties of the
support and catalysts obtained from the adsorption–desorption
isotherms of nitrogen. From the textural data, one can con-
clude that the cobalt loading causes a decrease in the specific
surface area of the monometallic catalyst by blocking the meso-
pores. This is due to the formation of cobalt spinel particles
resulting from calcination of the catalyst at 550 ◦C, as shown
in Fig. 1. The iridium loading does not affect the textural prop-
erties of the monometallic catalyst in spite of the XRD data
showing intense diffraction peaks for IrO2 [38,43] (Fig. 1), at
2θ angles of 28◦, 34.3◦, 40.1◦ and 53.8◦ corresponding to the
1 1 0, 1 0 1, 2 0 0 and 2 1 1 planes of IrO2. The textural prop-
erties of bimetallic catalysts show also a moderate decrease as
the iridium loading is increased. However, the XRD patterns
indicate that the diffraction peaks of IrO2 are only observed
for an iridium loading corresponding to an Co/Ir atomic ratio
lower than 20. Furthermore, the diffraction peaks characteris-
tic of Co3O4 appear, corresponding to the 2 2 0, 3 1 1 and 4 0 0
planes. The presence of iridium also provokes a slight broaden-
ing of the Co3O4 signals indicating a smaller size of the cobalt
particles on the support for the bimetallic catalysts than those of
monometallic ones.
Fig. 2. Core level Co 2p XPS spectra of (a) fresh Co-16, (b) fresh CoIr60, (c)
fresh CoIr30, (d) fresh CoIr20 and (e) fresh CoIr10 catalysts.
ies [44,45]. The TPR-H2 plots of monometallic and bimetallic
catalysts are shown in Fig. 2. It can be observed that the Co-
16 catalyst (Fig. 2A) shows two peaks of H2 consumption; one
peak centred at 324 ◦C with a shoulder at 294 ◦C and a broad
band from 382 to 730 ◦C. The peak at 324 ◦C with the shoulder
at 294 ◦C would correspond to the reduction of Co3O4 to metal-
lic cobalt whereas the broad band at high temperatures might
be ascribed to CoO interacting with the support with a different
intensity [46] or to the presence of particles of different sizes
[44]. However, the formation of cobalt silicate during calcina-
tion or the TPR test [46] by reaction of highly dispersed CoO
with the silica cannot be ruled out.
The monometallic iridium catalyst shows a TPR-H2 profile
(Fig. 2B), in which the presence of a very small and negative
peak at 90 ◦C can be observed due to a small desorption of hydro-
gen from the metallic iridium, formed during calcination, and
a broadened peak of hydrogen consumption at 220 ◦C. There is
also another small peak at 290 ◦C. These peaks of hydrogen con-
sumption either result from the reduction of IrOx particles which
are interacting to a different extent with the support, or IrOx par-
ticles trapped in the pores of the support delay their reduction by
In general TPR-H2 of cobalt supported catalysts show that the
two-step reduction of Co3O4 (via CoO) to cobalt metal can be
seen either as a single peak or two resolved peaks in TPR stud-

R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
129
hydrogen. This point was noticed by Reyes et al. [47] together
with the presence of several reduction peaks for iridium sup-
ported catalysts depending on the support used. Carnevillier et
al. [48] claimed that the reduction of 0.6 wt.% Ir supported on
Al2O3, took place in two peaks; the reduction of large Ir par-
ticles at low temperatures, whereas the well-dispersed iridium
particles were reduced at higher temperatures.
alysts. This value is between 15 eV for Co3O4 and 15.9 eV for
CoO [54,56–58]. These features seem to indicate the presence
of both Co3O4 and Co2+ species dispersed over the support, the
presence of Co2+ being higher for the CoIr60 and CoIr30 cata-
lysts. Therefore, it would seem that the presence of iridium ions
could impede the oxidation of Co2+ to form cobalt spinel during
the calcination process. For the atomic ratios obtained by XPS
(Table 2), the Co/Si ratio is increased with the iridium loading
up to the CoIr30 catalyst which was also the catalyst with the
highest presence of cobalt species, as observed above.
The XPS spectra in the iridium region have also been
recorded. Fig. 4A corresponds to the spectrum of CoIr30, where
the features of IrO2 are observed (4f7/2 ca. 62.0 eV) [47,59,60] as
was detected by XRD patterns, and a very low intensity doublet
with a 4f7/2 at 61.0 assigned to Ir0. A Co3p photoemission also
occurs in the region at 60.4 eV. The spectrum for Ir-2 catalyst
shows features characteristic of Ir0 (60.9 eV) and Ir4+ (61.7 eV)
The bimetallic catalysts show quite different TPR patterns to
the monometallic ones (Fig. 2C). Firstly, the small peak of IrO2
reduction in the single iridium catalyst appears now a with very
high intensity, meaning that the reduction of Ir(IV) is simulta-
neous to that of a fraction of cobalt. Therefore, the addition of
iridium to the cobalt supported catalyst leads to the cobalt oxides
undergoing reduction at lower temperatures. Thus, the reduction
peak appearing at 324 ◦C for the Co-16 catalyst shifts towards
lower temperatures (150–200 ◦C) in mixed oxides. Moreover,
theintensityofthispeakincreasesdependingontheCo/Iratomic
ratio, suggesting that the reduction of Co3O4 is favoured by the
presence of iridium. This means that cobalt and iridium ions are
present in the vicinity and influence the reducibility of the cobalt
spinel. However, the formation of lower cluster size of Co3O4
spinel due to the presence of iridium could also be responsible
of this better reducibility of cobalt species. At the same time,
the broad band at high temperatures of the Co-16 catalyst is nar-
rowed and ranges between 200 and 450 ◦C. This peak is assigned
to the reduction of Co2+ to Co0, indicating that the reduction of
the species interacting more strongly with the support is also
favoured by the presence of iridium. The promoter effect of
some metals [44,49–51] on the reducibility of cobalt oxides has
been reported previously by other authors. For example, Pt, Ru
and Re increase the reducibility of CoOx, although this phe-
nomena is not accompanied by an increased dispersion of CoOx
particles.
The XPS technique is a useful tool for identifying the oxi-
dation state of a dispersed metal oxide. XPS has been widely
used to determine the presence of either Co2+ or Co3O4 over
supported cobalt catalysts [46,52,53]. Owing to the small BE
difference and band broadening of the Co 2p3/2 signal, the posi-
tion of the Co 2p band alone proved unreliable for detecting the
unambiguous presence of Co2+ or Co3+. However, the Co2+ ion
may be distinguished by the presence of a shake-up satellite and
its separation from the photoelectronic band [54–56]. Moreover,
the doublet separation between the 2p3/2 and 2p1/2 signals may
be a parameter which could indicate the presence of Co2+ or
Co3+ [54,57].
The most valuable data from the XPS spectra are shown in
Table 2. The Co-16 catalyst shows a BE corresponding 2p3/2 of
779.5 eV as well as a low intensity shake-up satellite at 786.5 eV
(Fig. 3). These features are characteristic of the presence of
Co2+/Co3+ in the Co3O4 [46]. As the iridium is added to the
Co16 catalyst, a shift is observed of the 2p3/2 towards a higher
BE, however, this shift is unaffected by iridium loading (Fig. 3).
Moreover, the intensity of the shake-up satellite is increased,
reaching the highest value for the CoIr60 and CoIr30 catalysts
and it then decreases for higher iridium loading. The doublet sep-
aration between the 2p3/2 and 2p1/2 is also increased with iridium
loading by as much as 15.5 eV for the CoIr60 and CoIr30 cat-
Fig. 3. (A) H2-TPR curves of Co-16. (B) H2-TPR curves of Ir-2. (C) H2-TPR
curves of (a) CoPt60, (b) CoIr30, (c) CoIr20 and (d) CoIr10.

130
R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
Table 2
XPS data of CoIr catalysts
Catalyst
Co 2p3/2
Co 2p1/2
ꢀE2p
Isat (%)
Co/Si
Ir/Co
Ir/Si
S (%)
Co-16
Fresh
SO2
779.5
781.0
794.5
796.2
15.0
15.2
20.34
31.88
0.066
0.091
–
–
–
–
–
0.34
Ir-2
Fresh
SO2
–
–
–
–
–
–
–
–
–
–
–
–
0.006
0.003
–
0.40
CoIr60
Fresh
SO2
780.9
781.5
796.5
797.0
15.6
15.5
30.01
n.c.
0.065
0.075
0.067
0.010
0.004
0.005
–
0.48
CoIr30
Fresh
SO2
780.7
782.1
796.2
797.5
15.5
15.4
30.45
36.06
0.086
0.070
0.106
0.104
0.007
0.005
–
0.57
CoIr20
Fresh
SO2
780.9
781.4
796.0
796.6
15.1
15.2
28.78
33.54
0.047
0.077
0.087
0.135
0.006
0.007
–
0.45
CoIr10
Fresh
SO2
780.9
781.4
796.1
796.9
15.2
15.5
27.79
31.99
0.051
0.086
0.157
0.223
0.007
0.015
–
0.49
(Fig. 4B). These results are in accordance with the XRD patterns,
which only show the reflection peaks corresponding to IrO2, and
with TPR analysis, which in this case showed a negative peak
ascribed to H2 desorption from Ir0.
The TPD-NO profiles of the CoIr catalysts as well as of the
Co-16 and Ir-2 catalysts are shown in Fig. 5. For the TPD-NO
of the Co-16 catalyst (Fig. 5A), the TPD-NO plot shows three
peaks at 150, 220 and 373 ◦C, the intensity of the 150 and 220 ◦C
peaks being almost the same. Furthermore, the only nitrogenous
species detected was NO together with a very small amount of
N2O at higher temperatures. These results are similar to those
reported by Haneda et al. [62] for Co3O4 which showed a very
broad NO desorption band from 100 to 300 ◦C with no desorp-
tion of N2 or O2, suggesting that the NO decomposition did not
take place over Co3O4. Much research on the FTIR study of NO
adsorption onto cobalt zeolites has been reported [62–67]. On
the basis of these reports we may assign the two peaks at lower
temperature to the desorption of NO from Co2+. The NO may
be adsorbed as Co2+-NO or Co2+-(NO)2. Although Lee et al.
[63] claimed that the dispersed Co2+ ions in the ZSM-5 zeolite
are responsible for NO adsorption instead of the cobalt oxide.
Zhu et al. [67], studying the NO adsorption over Co-ZSM-5 with
different exchanged-Co content, observed that at temperatures
above 300 ◦C a band appeared which was assigned to the adsorp-
tion of NO onto Co3O4 as the zeolite was over-exchanged and
cobalt oxides were present. Thus, the broad band at 373 ◦C may
be due to the NO desorption from the Co3O4 particles dispersed
over the support.
The TPD-NO of the Ir-2 catalyst (Fig. 5A) calcined at 550 ◦C
shows a single broad band of NO desorption extending from 50
to 250 ◦C, indicating the existence of different interactions of
NO molecules with the particles of IrO2.
The amount of NO desorbed from the bimetallic catalysts
(Fig. 5B) is considerably higher than from monometallic ones
Fig. 4. Core level Ir 4f XPS spectra of (A) fresh CoIr30 and (B) fresh Ir-2.

R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
131
Fig. 5. (A) TPD curves of NO adsorbed on Co-16 (ꢀ) and Ir-2 (ꢁ). (B) TPD
curves of NO adsorbed on CoIr60 (ꢂ), CoIr30 (×), CoIr20 (ꢀ) and CoIr10 (᭹).
in spite of all catalysts having the same cobalt loading, indicat-
ing extensive involvement of iridium on the adsorption of this
molecule. Moreover, the sum of the individual amounts of des-
orbed NO from the monometallic catalysts does not coincide
with the NO desorbed from the bimetallic catalyst and hence,
there are more active sites able to adsorb NO than is the case with
the monometallic catalysts, demonstrating that the presence of
iridium improves the cobalt oxide dispersion over the support.
This fact has been established by other authors in the case of
cobalt catalysts promoted with Pd [63], alkali metals [61] or Pt
[28]. CoIr60 and CoIr30, having the maximum superficial con-
centration of Co2+ ions, retain the least amount of NO, which is
also desorbed at lower temperatures.
Fig. 6. (A) NO conversion as a function of reaction temperature for CoIr60 (×),
CoIr30 (ꢂ), CoIr20 (ꢃ) and CoIr10 (ꢀ) catalysts. Experimental conditions: NO
(1000 ppm), NH3 (1000 ppm), O2 (2.5 vol.%); total flow rate: 42 cm3 min−1
.
(B) NO conversion as a function of reaction temperature for Co-16 (ꢀ) and Ir-2
) catalysts. Experimental conditions: NO (1000 ppm), NH3 (1000 ppm), O2
(2.5 vol.%); total flow rate: 42 cm3 min−1
(
.
tant as described above. Moreover, there are other competitive
reactions in which ammonia is oxidised by the oxygen to yield
N2, N2O or NO.
These reactions are:
4NH3 + 3O2 → 2N2 + 6H2O
2NH3 + 2O2 → N2O + 3H2O
4NH3 + 5O2 → 4NO + 6H2O
(III)
(IV)
(V)
3.2. Catalytic performances
All authors currently agree that the SCR of NO with NH3 in
the presence of O2 is based on the following reaction:
These reactions lead to ammonia consumption and therefore
they are undesirable since the reductant is consumed by oxygen
instead of NO.
The synthesised catalysts have been tested in the SCR of NO
with ammonia in the presence of excess oxygen. Fig. 6A and
B show the NO conversion of both bimetallic and monometal-
lic catalysts in the absence of water and SO2 in the feed. All
the bimetallic catalysts (Fig. 6A) show a very similar catalytic
behaviour with high activity at low temperature, with the CoIr30
catalyst reaching the highest NO conversion (98%) between 170
4NO + 4NH3 + O2 → 4N2 + 6H2O
(I)
Thus, the SCR selectivity of NO with NH3 must lead to the
formation of N2 instead of N2O which is produced by the non-
selective reaction:
4NO + 4NH3 + 3O2 → 4N2O + 6H2O
(II)
Reactions (I) and (II) lead to the removal of NO, although
reaction (II) must be avoided because N2O is a harmful pollu-

132
R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
and 190 ◦C. Furthermore, for the rest of the bimetallic catalysts,
the NO conversion is higher than 90%, even at 150 ◦C for the
CoIr60 catalyst. The monometallic catalysts (Fig. 6B) show a
lower NO conversion than the bimetallic ones, thus the Co-16
catalyst shows a high NO conversion at 150 ◦C (87%), whereas
the Ir-2 catalyst only shows a 59.2% of NO conversion at 200 ◦C.
Therefore, the addition of iridium to the Co-16 catalyst seems
to be beneficial for the NO conversion with ammonia. This fact
may be related to the better dispersion of the cobalt over the
support as the characterisation results have shown previously.
Thus, H2-TPR revealed an easy reducibility of Co3O4 species
due to their higher dispersion in the presence of iridium, as
confirmed by XRD data. On the other hand, the catalytic perfor-
mance of supported iridium catalysts is quite different when CO
or hydrocarbons are used as reductant agents. Thus, Nakatsuji
[68] reported NO conversions as high as 90% using propene as
a reductant for 2 wt.% Ir supported on titania.
The NO conversion for each catalyst as a function of the reac-
tion temperature show a volcano curve plot. It can be observed
that the shape of the curve of the catalyst with the lowest iridium
loading (the CoIr60 catalyst) is quite similar to the monometal-
lic Co-16 catalyst although with better NO conversion, whereas
the catalysts with the highest iridium loadings, the CoIr20 and
CoIr10catalysts, displaycurveswithashapelikeIr-2, butalways
with a better coversion of NO. Finally, the catalyst with an
intermediate iridium loading, the CoIr30 catalysts, give a dif-
ferent curve with a very wide temperature window of operation
(120–320 ◦C).
ratio, i.e. with similar cobalt dispersion over the support and
hence a predictably similar catalytic performance. On the basis
of the cobalt dispersion, the CoIr30 catalyst with the highest
Co/Si atomic ratio value exhibits the greatest NO conversion.
As regards the concentration of active centres involved in this
reaction, it is again the CoIr30 which has the maximum amount
of Co2+ ions as was shown from XPS analysis. In a previous
study [40], we established that the formation of a more crys-
talline Co3O4 spinel phase on the catalysts leads to a decrease
in the NO conversion. As a consequence, the presence of well
dispersed Co2+ ions leads to the transference of oxygen from
the gas phase to the catalyst and oxidation of Co2+ to Co3+,
where NO is adsorbed and oxidised to NO2. The NO2 formed
is then reduced to N2 by ammonia. This result is in agree-
ment with the mechanism proposed by other authors on the
SCR of NO over cobalt catalysts, where the presence of CoO
is necessary to catalyse the oxidation of NO to NO2, consid-
ered as the crucial step in the performance at low temperatures
[69].
Table 3 shows the catalytic data for each of the catalysts
tested. There are some notable features in these results. Firstly,
the ammonia is entirely consumed; this observation is notable
since if the NO conversion is not 100%, then some ammonia
is consumed in the non-selective reactions (reactions (III) and
(IV)). This point is very important because no ammonia escapes
to the atmosphere and so is not emitted as a pollutant to the
environment. Secondly, the selectivity to nitrogen is higher than
92% demonstrating that reaction (I) is predominant. The excep-
tion is the Ir-2 catalyst which produces a considerable amount
of N2O and a very low selectivity to N2, therefore making the
catalyst unsuitable for the SCR of NO with ammonia. Thus,
according to the catalytic data found so far, these catalysts could
be candidates for the SCR of NO with ammonia and of these,
CoIr30 would be the best. These results are promising because
The catalytic performances of this set of mixed catalysts
might be related to the Co/Si atomic ratio values obtained by
XPS studies. Table 2 shows the Co/Si atomic ratio for the both
monometallic and bimetallic catalysts. From these data it is pos-
sible to consider Co-16 and CoIr60 as one group of catalysts and
CoIr20 and CoIr10 as another, having the same Co/Si atomic
Table 3
Catalytic data
Catalyst
Temperaturea (◦C)
NO conversion (%)
NH3 conversion (%)
N2O formation (ppm)
N2 selectivity (%)
Co-16
150
200
150
180
200
190
87.4
59.2
96.2
98.5
96.5
91.4
100
93.5
99.6
100.0
100.0
100.0
69.0
530.4
55.6
45.7
35.5
65.0
92.4
2.9
94.5
95.3
96.3
97.7
Ir − 2
CoIr60
CoIr30
CoIr20
CoIr10
Dry conditions
10% (v/v) H2O
Co-16
200
250
200
190
170
170
83.8
51.7
83.2
90.3
51.7
81.6
83.0
49.3
100.0
100.0
50.8
30.1
370.5
76.6
32.2
88.9
96.3
31.7
91.0
96.3
83.4
86.2
Ir − 2
CoIr60
CoIr30
CoIr20
CoIr10
77.8
104.9
Co-16
220
220
210
200
190
210
31.2
8.2
29.4
27.3
40.5
49.4
32.3
12.8
30.9
28.2
39.3
55.5
10.3
105.0
30.2
7.9
173.4
186.2
96.8
0.0
89.6
97.1
55.8
62.1
Ir − 2
CoIr60
CoIr30
CoIr20
CoIr10
100 ppm SO2
a
Temperature at which the catalysts reach their highest conversion.

R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
133
they are better than those reported by Lee et al. [63] for a FER
exchanged with cobalt (1.47 wt.%) and iridium (0.51 wt.%)) and
tested in the SCR of NO with methane at a GHSV of 14.000 h−1
.
These results are also an improvement on those reported previ-
ously for cobalt and platinum supported on a zirconium doped
mesoporous silica [40].
When 10% (v/v) of water is added to the feed, surprisingly,
the catalytic performances do not undergo substantial changes.
In fact, the Co-16 catalyst exhibits only a small decrease in the
NO conversion of close to 3% and even improves the selectivity
to N2 (Table 3). Only the temperature of maximum conversion
is shifted to 200 ◦C. The Ir-2 catalyst also maintains most of
its catalytic activity, decreasing by only 7%, but improving the
selectivity to N2 (from 3 to 32%), although this catalyst still
shows a high level of N2O production.
In the presence of water, the bimetallic catalysts also suffer a
decrease in their NO conversions, especially the CoIr20 catalyst.
Amongthebimetalliccatalysts, theCoIr30catalystgivesthebest
performance, maintaining its NO conversion at about 90% with a
selectivity to N2 close to 100% and with an ammonia conversion
of 100%. It is known that water competes with NO and NH3 for
the active sites [25] on the catalyst and this is the main reason
for the loss of catalytic activity. So the CoIr30 catalyst with the
best dispersion is again seen to achieve the best performance.
The effect of SO2 in the feed has also been studied. If the SO2
concentration in the feed is 20 ppm, the CoIr10 catalyst gives a
NO conversion of 72% at 220 ◦C. The concentration of 20 ppm
of SO2 in the feed has been used by many authors to study the
influence of this molecule in the catalytic activity of iridium cat-
alysts and a promotional effect has been reported in the SCR
of NO with H2 or CO as reductant. However, when 100 ppm of
SO2 were fed to this set of catalysts, Fig. 7A and B, the activ-
ity of the catalysts is reduced, the CoIr10 catalyst giving the
highest NO conversion of 49% at 210 ◦C (Table 3). In contrast
to the Ir-2 sample, which exhibits the lowest values of conver-
sion, the catalysts with the higher cobalt loading are now the
most active. This means that IrO2 alone is not active under these
experimental conditions, and there is possibly a synergic effect
between Co/Ir leading to higher conversions for the CoIr10 and
CoIr20 catalysts, but in both cases the N2O formation becomes
quite significant. For an insight into the influence of SO2 in
the modification of the catalytic behaviour, the spent catalysts
were analysed by XPS in the Co 2p region. This peak shifted
to higher B.E. values and the intensity of the shake-up satellite
was increased, revealing a substantial increment in the superfi-
cial concentration of Co2+ ions originating from the reductant
properties of SO2, in spite of the reaction being run in excess
oxygen. On the other hand, the CoIr20 catalyst exhibits a new
peak in the Ir region corresponding to Ir0. Thus, the presence
of SO2 increases the superficial concentration of Co2+ ions and
impedes the formation of the Co2+/Co3+ pairs to such an extent
that there is a lower NO conversion. On the other hand, by XPS
is detected the presence of sulphur as SO42−in the spent cata-
lyst (Table 2), so the incorporation of SO2 to the feed leads to
sulfidation of the transition metal cation and hence it inhibits
both NO and NH3 adsorption as a previous step to catalysts
reaction.
Fig. 7. (A) NO conversion as a function of temperature for CoIr60 (×), CoIr30
(ꢂ), CoIr20 (ꢃ) and CoIr10 (ꢀ) catalysts in the presence of 100 ppm SO2.
Experimental conditions as mentioned in Fig. 6. (B) NO conversion as a function
of temperature for Co-16 (ꢀ) and Ir-2 ( ) catalysts in the presence of 100 ppm
SO2. Experimental conditions as mentioned in Fig. 6.
4. Conclusions
Bimetallic Co–Ir catalysts supported on zirconium meso-
porous silica are good catalysts for the SCR of NO by ammonia
in excess oxygen at low reaction temperatures, with a high selec-
tivity to N2 and a very low selectivity to N2O. The reductant
agent, NH3, was totally converted. The incorporation of iridium
has a beneficial effect promoting the reducibility and the disper-
sion of cobalt oxides and reducing the formation of the cobalt
spinel during the calcination process. The addition of water to
the feed does not produce significant changes in the catalytic
performance. However, the presence of 100 ppm pf SO2 in the
feed produces a dramatic decrease of the catalytic activity prob-
ably due to an increase of the superficial Co2+ concentration
originating from the reducing properties of SO2.
Acknowledgement
Financial support from the European Brite/EuRam Contract
BRPR-CT97-0545 is acknowledged with thanks.

134
R. Moreno-Tost et al. / Journal of Molecular Catalysis A: Chemical 248 (2006) 126–134
[39] T. Yoshinari, K. Sato, M. Haneda, Y. Kintaichi, H. Hamada, Catal.
References
Commun. 2 (2001) 155.
[40] R. Moreno-Tost, J. Santamar´ıa-Gonza´lez, P. Maireles-Torres, E.
Rodr´ıguez-Castello´n, A. Jime´nez-Lo´pez, Appl. Catal. B 38 (2002) 51.
[41] D.J. Jones, J. Jime´nez-Jime´nez, A. Jime´nez-Lo´pez, P. Maireles-Torres,
P. Olivera-Pastor, E. Rodr´ıguez-Castello´n, J. Rozie`re, Chem. Commun.
(1997) 431.
[1] R.V. Pouyat, M.A. McGlinch, Environ. Sci. Policy 1 (1998) 249.
[2] G. Busca, L. Lietti, G. Rams, F. Berti, Appl. Catal. B 18 (1998) 1.
[3] I.E. Wachs, G. Deo, B.M. Weckhuysen, A. Andreini, M.A. Vuurman,
M. De Boer, M.D. Amiridis, J. Catal. 161 (1996) 211.
[4] L. Lietti, Appl. Catal. B 10 (1996) 281.
[42] E. Rodr´ıguez-Castello´n, A. Jime´nez-Lo´pez, P. Maireles-Torres, D.J.
Jones, J. Rozie`re, M. Trombetta, G. Busca, M. Lenarda, L. Storaro,
J. Solid State Chem. 175 (2003) 159.
[43] M. Haneda, Y. Pusparatu, I. Kintaichi, M. Nakamura, T. Sasaki, H.
Fujitani, Hamada, J. Catal. 229 (2005) 197.
[44] D. Schanke, S. Vada, E.A. Blekkan, A.M. Hilmen, A. Hoff, A. Holmen,
J. Catal. 156 (1995) 85.
[45] S. Sun, N. Tsubaki, K. Fujimoto, Appl. Catal. A 202 (2000) 121.
[46] A. Mart´ınez, C. Lo´pez, F. Ma´rquez, I. D´ıaz, J. Catal. 220 (2003) 486.
[47] P. Reyes, M.C. Aguirre, G. Pecchi, J.L.G. Fierro, J. Mol. Catal. A 164
(2000) 245.
[48] C. Carnevillier, F. Epron, P. Marecot, Appl. Catal. A 275 (2004) 25.
[49] S.A. Hosseini, A. taeb, F. Feyzi, F. Yaripour, Catal. Commun. 5 (2004)
137.
[50] W.S. Epling, P.K. Cheekatamarla, A.M. Lane, Chem. Eng. J. 93 (2003)
61.
[51] G. Jacobs, T.K. Das, Y. Zhang, J. Li, G. Racoillet, B.H. Davis, Appl.
Catal. A 233 (2002) 263.
[52] Y. Okamoto, K. Nagata, T. Adachi, T. Imanaka, K. Inamura, T. Tayku,
J. Phys. Chem. 95 (1991) 310.
[53] A.Y. Khodakov, A. Griboval-Constant, R. Bechara, V.L. Zholobenko, J.
Catal. 206 (2002) 230.
[54] D. Pietrogiacomi, S. Tuti, M.C. Campa, V. Indovina, Appl. Catal. B 28
(2000) 43.
[55] G. Fierro, M.A. Eberhardt, M. Houalla, D.M. Hercules, W.K. Hall, J.
Phys. Chem. 100 (1996) 8468.
[56] A. Frydman, D.G. Castner, M. Schmal, C.T. Campbell, J. Catal. 152
(1995) 164.
[57] N. Li, A. Wang, M. Zheng, X. Wang, R. Cheng, T. Zhang, J. Catal. 225
(2004) 307.
[58] Y. Brik, M. Kacimi, M. Ziyad, F. Bozon-Verduraz, J. Catal. 202 (2001)
118.
[59] H. Tian, T. Zhang, X. Sun, D. Liang, L. Li, Appl. Catal. A 210 (2001)
55.
[60] K. Nakagawa, N. Ikenaga, T. Suzuki, T. Kobayashi, M. Haruta, Appl.
Catal. A 169 (1998) 281.
[5] M. Meng, P.-Y. Lin, Y.-L. Fu, Catal. Lett. 48 (1997) 213.
[6] M. Antunes Pereira da Silva, M. Schmal, Catal. Today 85 (2003) 31.
[7] Z. Chajar, V. Le Chanu, M. Primet, H. Praliaud, Catal. Lett. 52 (1998)
97.
[8] H.K. Shin, H. Hirabayashi, H. Yahiro, M. Watanabe, M. Iwamoto, Catal.
Today 26 (1995) 13.
[9] A.D. Cowan, R. Du¨mpelmann, N.W. Cant, J. Catal. 151 (1995) 356.
[10] H.Y. Chen, T. Voskoboinikov, W.M.H. Sachtler, Catal. Today 54 (1999)
483.
[11] B.K. Cho, J.E. Yie, Appl. Catal. B 10 (1996) 263.
[12] W. Wang, S.J. Hwang, Appl. Catal. B 5 (1995) 187.
[13] R.Q. Long, R.T. Yang, Catal. Lett. 74 (2001) 201.
[14] P. Fabrizioli, T. Bu¨rgi, A. Baiker, J. Catal. 206 (2002) 143.
[15] W.S. Kijlstra, D.S. Brands, E.K. Poels, A. Bliek, Catal. Today 50 (1999)
133.
[16] G. Centi, S. Perathoner, D. Biglino, E. Giamello, J. Catal. 151 (1995)
75.
[17] G. Delahay, B. Coq, S. Kieger, B. Neveu, Catal. Today 54 (1999) 431.
[18] B.L. Duffy, H.E. Curry-Hyde, N.W. Cant, P.F. Nelson, J. Catal. 154
(1995) 107.
[19] H. Schneider, M. Maciejewski, K. Ko¨hler, A. Wokaun, A. Baiker, J.
Catal. 157 (1995) 312.
[20] L. Chmielarz, R. Dziembaj, T. Grybek, J. Klinik, T. Lojewski, D.
Olszewska, H. Papp, Catal. Lett. 68 (2000) 95.
[21] Y. Li, J.N. Armor, Appl. Catal. B 13 (1992) L31.
[22] Y. Li, J.N. Armor, J. Catal. 145 (1994) 1.
[23] H. Ohtsuka, Appl. Catal. B 33 (2001) 325.
[24] Y.K. Park, J.W. Lee, C.W. Lee, S.E. Park, J. Mol. Catal. A 158 (2000)
173.
[25] Y. Li, P.J. Battavio, J.N. Armor, J. Catal. 142 (1993) 561.
[26] P. Budi, R.F. Howe, Catal. Today 38 (1997) 175.
[27] B. Wen, J. Jia, S. Li, T. Liu, L.X. Chen, W.M.H. Sachtler, Phys. Chem.
Chem. Phys. 4 (2002) 1983.
[28] L. Gutierrez, A. Boix, J.O. Petunchi, J. Catal. 179 (1998) 179.
[29] D.A. Pen˜a, B.S. Uphade, P.G. Smirniotis, J. Catal. 221 (2004) 421.
[30] M. Brandhorst, J. Zajac, D.J. Jones, J. Rozie`re, M. Womes, A. Jime´nez-
Lo´pez, E. Rodriguez Castello´n, Appl. Catal. B 55 (2005) 267.
[31] L. Chmierlarz, P. Kustrowski, A.R. Lasocha, D. Majda, R. Dziembaj,
Appl. Catal. B 35 (2002) 195.
[32] L. Chmielarz, P. Kustrowski, M. Zbroja, B. Gil-Kanp, J. Datka, R.
Dziembaj, Appl. Catal. B 45 (2003) 103.
[33] X. Wang, H.-Y. Chen, W.M.H. Sachtler, Appl. Catal. B. 26 (2000) L227.
[34] C. Wo¨gerbauer, M. Maciejewski, A. Baiker, Appl. Catal. B 34 (2001)
11.
[35] C. Wo¨gerbauer, M. Maciejewski, A. Baiker, U. Go¨bel, Topics Catal.
16–17 (2001) 181.
[36] M. Ogura, A. Kawamura, M. Matsukata, E. Kikuchi, Chem. Lett. (2000)
146.
[61] L.A. Silva, V.A. Alves, S.C. de Castro, J.F.C. Boodts, Colloids Surf. A:
Physicochem. Eng. Aspects 170 (2000) 119.
[62] M. Haneda, Y. Kintaichi, N. Bion, H. Hamada, Appl. Catal. B 46 (2003)
473.
[63] C.W. Lee, P.J. Chong, Y.C. Lee, C.S. Chin, L. Kewan, Micropor. Mater.
12 (1997) 21.
[64] T.J. Lee, I.-S. Nam, S.-W. Ham, Y.-S. Baek, K.-H. Shin, Appl. Catal. B
41 (2003) 115.
[65] K. Hadjiivanov, B. Tsyntsarski, T. Nikolova, Phys. Chem. Chem. Phys.
1 (1999) 4521.
[66] F. Geobaldo, B. Onida, P. Rivolo, F. Di Renzo, F. Fajula, E. Garrone,
Catal. Today 70 (2001) 107.
[67] C.Y. Zhu, C.W. Lee, P.J. Chong, Zeolites 17 (1996) 483.
[68] T. Nakatsuji, Appl. Catal. B 25 (2000) 163.
[69] A.Y. Stakheev, C.W. Lee, S.J. Park, P.J. Chong, Catal. Lett. 38 (1996)
327.
[37] A. Wang, D. Liang, C. Xu, X. Sun, T. Zhang, Appl. Catal. B 32 (2001)
205.
[38] T. Yoshinari, K. Sato, M. Haneda, Y. Kintaichi, H. Hamada, Appl. Catal.
B 41 (2003) 157.