Nitric oxide reduction catalyzed by amorphous copper-zirconium and copper-titanium alloys

Article abstract of DOI:10.1016/S0925-8388(00)01393-1

Amorphous alloys have been regarded as interesting from the catalytic point of view because they lack of long range ordering. Moreover, the surface of amorphous alloys is rich in low-coordination sites and defects which play an important role in catalysis. Amorphous Cu-Zr and Cu-Ti ribbons were produced by melt spinning methods. The surface of amorphous materials was prepared, before catalytic tests, by immersion in HF solutions or by H₂ flow reduction. A selective dissolution of first transition metals was exhibited by HF surface pre-treatment. A porous and Cu-rich surface was obtained. In the reduction of NO_x with propylene in presence of O₂, amorphous ribbons exhibited an initial high activity that decreased until the steady-state was reached. This behaviour can be explained with the oxidation of copper active sites by the oxygen contained in the reaction stream. Cu-Zr alloys showed higher catalytic activity than Cu-Ti alloys. Moreover, the HF treatment increased the activity of the materials, and the HF treated alloys were more active than the H₂ treated alloys.

Full text of DOI:10.1016/S0925-8388(00)01393-1

Journal of Alloys and Compounds 317–318 (2001) 590–594
L
www.elsevier.com/locate/jallcom
Nitric oxide reduction catalyzed by amorphous copper–zirconium and
copper–titanium alloys
a
b
a ,
*
M. Dabal a` , P. Canu , M. Magrini
a
Dipartimento di Innovazione Meccanica e Gestionale, Universit a` di Padova, Via Marzolo 9, 35100 Padova, Italy
b
Istituto di Impianti Chimici, Universit a` di Padova, Via Marzolo 9, 35100 Padova, Italy
Abstract
Amorphous alloys have been regarded as interesting from the catalytic point of view because they lack of long range ordering.
Moreover, the surface of amorphous alloys is rich in low-coordination sites and defects which play an important role in catalysis.
Amorphous Cu–Zr and Cu–Ti ribbons were produced by melt spinning methods. The surface of amorphous materials was prepared,
before catalytic tests, by immersion in HF solutions or by H2 flow reduction. A selective dissolution of first transition metals was
exhibited by HF surface pre-treatment. A porous and Cu-rich surface was obtained. In the reduction of NO with propylene in presence of
x
O , amorphous ribbons exhibited an initial high activity that decreased until the steady-state was reached. This behaviour can be explained
2
with the oxidation of copper active sites by the oxygen contained in the reaction stream. Cu–Zr alloys showed higher catalytic activity
than Cu–Ti alloys. Moreover, the HF treatment increased the activity of the materials, and the HF treated alloys were more active than the
H treated alloys.
 2001 Elsevier Science B .V . All rights reserved.
2
Keywords: Amorphous alloys; Activation surface treatments; Nitric oxide reduction; Propylene; Catalyst
1
. Introduction
framework as in supported catalysts [2]. Since amorphous
materials can be prepared in a broad range of compositions
relatively free from thermodynamic constraints, the possi-
bility of developing special catalysts of high selectivity is
large [3]. Unfortunately, the amorphous alloys prepared by
liquid quenching do not show high catalytic activity,
because the surface of these materials is covered by an
oxyhydroxide layer [4]. For this reason glassy alloys are
not directly used, but some surface treatments are applied
for the enhancement of the catalytic activity.
In the past years amorphous alloys were the subject of
many studies [5–13]. Most of these investigations were
carried out on zirconium and titanium materials, especially
alloyed with copper. In these catalysts the active metal is
copper and the activation treatments are made to obtain a
surface copper enrichment. In oxidant environment of
reaction, the most effective activation treatments to en-
hance the catalytic activity of Cu–Ti and Cu–Zr alloys are
carried out both by immersion in HF solutions and by H2
attack at reaction temperature [14].
The development of melt quenching methods allowed
metallic glasses to become available from the beginning of
the of 1970s, and these metastable materials have attracted
the attention of researchers because of their exceptional
physical and chemical properties. Since polycrystalline
bimetallic catalysts are widely used, and there is a strong
structure dependence of the reaction rate in many re-
actions, the use of amorphous alloys in catalytic syntheses
has received considerable attention [1]. In fact, these
materials are thought to consist ideally of metal clusters
similar to the ones employed in conventional catalysts.
However, they are of uniform size, self-supported and
placed in a homogeneous chemical environment. The
metal ribbon exhibits no porosity and the active surface is
very similar to the geometrical surface. Then, the reagent’s
diffusion towards the active sites is not hindered by porous
*
Corresponding author. Tel.: 139-049-827-5503; fax: 139-049-827-
500.
E-mail address: mmagrini@ux1.unipd.it (M. Magrini).
In this work the effect of activation treatments in HF or
5
in H on Cu–Zr and Cu–Ti amorphous alloys structures
2
0
925-8388/01/$ – see front matter
 2001 Elsevier Science B .V . All rights reserved.
PII: S0925-8388(00)01393-1

M. Dabal a` et al. / Journal of Alloys and Compounds 317–318 (2001) 590–594
591
and their catalytic performance in the NO reduction with
x
propylene in presence of O is reported and discussed.
2
2
. Experimental details
One mm wide, 30 mm thick Cu Zr , Cu Ti and
6
0
40
60
40
Cu Zr ribbons were produced by melt spinning method
4
0
60
2
1
in vacuum at a cooling wheel rate of 3000 cm s . The
catalytic tests were carried out in a flow-type microreactor
Fig. 1) with about 0.6 g alloy and GHSV 26 550 h . The
2
1
(
reaction products (N2 and NO) were subjected to GC
analysis by a HP 6890 Plus gas chromatograph with
thermal conductivity detector, a 6 ft length molecular sieve
5
A column, and He as carrier gas. The gas reactant
moisture was composed of 630 ppmv of NO , 1370 ppmv
x
of C H and 7.7% vol. of O . The temperature of reaction
3
6
2
was 2608C.
The amorphous alloys underwent two different activa-
tion treatments: one with immersion in 1 M HF solution
for 1 min, washed with H O and still wet put into reactor;
2
one with H stream at 2308C for 4 h into the reactor. The
2
structure of the alloys was studied by X-ray diffraction
Fig. 2. X-ray patterns of the as-quenched amorphous materials.
(
XRD). X-ray diffraction patterns were collected with a
Bragg–Brentano powder diffractometer using Cu Ka
˚
radiation (l51.54178 A) and a graphite monochromator in
The first-neighbour distance a, evaluated by the Bragg
the diffracted beam. The surface of the alloys was investi-
gated by scanning electron microscopy (SEM) with a
Cambridge Stereoscan 440 microscope equipped with
EDX microbeam.
relation, is multiplied by a correction factor of 1.23 [15]
1.23g
a 5
2 sin ub
and the results are summarized in Table 1.
3
. Results and discussion
The results are in agreement with data obtained with
XRD structure factors and/or neutron scattering diffraction
factors reported in literature [16].
The amorphous condition of the as-quenched materials
was checked by X-ray diffractograms. Only broad peaks
characteristic of an amorphous structure are observed (Fig.
The surfaces of the two sides of the as-quenched alloys
show the characteristics of metallic glasses prepared by
melt spinning methods (Fig. 3). The inner sides show
small, parallel grooves due to contact with the cooling
wheel. The outer sides exhibit large randomly located
hillocks.
2
).
Fig. 4 summarizes the catalytic behaviour of Cu Zr
6
0
40
catalyst after different pretreatments. Note that the reactor
is continuous and the figure reports the activity variations
vs. time.
The HF treated material exhibits a high catalytic activity
which decreases in the course of the reaction until it
reaches stable and low activity after 150 min. The catalytic
performance of H treated material is weaker than the HF
2
Table 1
X-ray peaks angle and first neighbour distance a of as quenched alloys
Alloy
ub (degree)
a (nm)
Cu60Ti40
Cu60Zr40
Cu40Zr60
42.5
40.6
37.4
0.2615
0.2732
0.2957
Fig. 1. Catalytic plant for the reduction of NOx with propylene: (1) is
C H in He gas cylinder; (2) is O2 in He gas cylinder; (3) is NO in He
gas cylinder; (a) is the reactor and (b) is the water cooling of products.
3
6
x

5
92
M. Dabal a` et al. / Journal of Alloys and Compounds 317–318 (2001) 590–594
Fig. 5. SEM image of Cu60Zr40 catalyst surface treated with H2.
particles (Fig. 5). These clusters are active in NO reduc-
x
tion, but the O contained in the reaction stream oxidizes
2
the active sites of the surface and reduces the catalytic
activity very quickly.
However, as Fig. 6 shows, the oxidation of copper active
sites is partially reversible. In fact, a 2 h hydrogen
treatment partially restores the activity of the catalyst.
Unfortunately, this restoration process is less and less
effective with the reaction process.
Since Zr/Zr oxides dissolve in HF more readily than
Cu, HF treatment gives drastic changes in the surface
morphology. The HF treatment produces a higher porous
surface than the H treatment does, with a high depletion
2
Fig. 3. SEM images of the as-quenched Cu40Zr60 amorphous alloy: (a)
outer side; (b) inner side.
in Zr content, as it can be seen in Fig. 7. These surfaces
show higher activity, but the oxidation, due to the reaction
stream, produces the same activity depletion observed for
treated alloy, but the activity vs. time curves show the
same value at 150 min. After this period the activity of H2
treated catalyst decreases to zero. This may be interpreted
by the SEM investigation.
the H treated catalyst.
2
However, the deeper porosity of HF treated surfaces
keeps a residual low activity for a long time. The catalytic
performance of the Cu Ti alloy after different pretreat-
6
0
40
The H treatment causes a devitrification of the amor-
ment is reported in Fig. 8. The H treated sample shows a
2
2
phous surface and produces an enrichment of small copper
Fig. 6. Conversion of NO vs. time of Cu60Zr40 catalyst pretreated with
x
Fig. 4. Time-on-stream curves of Cu60Zr40 catalysts pretreated by HF
and H2.
H : after 270 and 390 min the catalyst underwent an H treatment for 2 h
at reaction temperature.
2
2

M. Dabal a` et al. / Journal of Alloys and Compounds 317–318 (2001) 590–594
593
Fig. 9. SEM image of Cu60Ti40 catalyst surface treated with H2.
the oxide layer. The low activity is probably due to a small
amount of metallic copper on the inner ribbon’s surface
[17].
Moreover, because the Ti oxides are less reactive than
Zr oxides with HF solutions, the HF treated surfaces show
lower porosity than the Zr alloys (Fig. 10). Consequently,
a significantly smaller copper exposure occurs and the
catalytic activity is lower [11].
In fact, as summarized in Fig. 11, the Cu–Ti catalyst
exhibits a lower reactivity than Cu–Zr alloys. However,
after 150 min, the catalytic activity of the three materials
was found to be about the same. This indicates that the
three glassy metals have the same kind of copper active
sites, but in different number, in agreement with the SEM
observations.
Fig. 7. SEM image of Cu60Zr40 catalyst surface treated with HF: (a) top
view; (b) lateral view.
very low activity for few minutes, while the HF activated
material exhibits higher activity for more time. The H2
activation is not successful in the case of Cu–Ti alloy. In
4
. Conclusions
fact, the H treated Cu–Ti surfaces do not exhibit forma-
tion of copper clusters as the Cu–Zr surfaces do (Fig. 9).
This depends on the fact that the Ti oxides are more
2
Copper containing alloys show catalytic activity in the
reduction of NO with propylene in presence of O only
x
2
after activation treatments. However, the O of the reaction
2
stable than Zr oxides and the H is ineffective to reduce
2
Fig. 8. Time-on-stream curves of Cu60Ti40 catalysts pretreated by HF and
H2.
Fig. 10. SEM image of Cu60Ti40 catalyst surface HF treated.

5
94
M. Dabal a` et al. / Journal of Alloys and Compounds 317–318 (2001) 590–594
9
803113214 002, Project Alloys and Intermetallic com-
]
pounds: thermodynamics, physical properties, reactivity),
and Prof. Baricco of University of Turin (Italy) for
supplying of amorphous alloys.
References
[
[
1] A. Molnar, G .V . Smith, M. Bartok, Adv. Catal. 36 (1989) 329.
2] R. Schloegl, in: S. Steeb, H. Warlimont (Eds.), Rapidly Quenched
Metals, Vol. 2, Elsevier Publishers, 1985, p. 1723.
[
3] D.L. Cocke, C. Yoon, in: S. Steeb, H. Warlimont (Eds.), Rapidly
Quenched Metals, Vol. 2, Elsevier Publishers, 1985, p. 1497.
4] D. Gasser, A. Baiker, Appl. Catal. 48 (1989) 279.
[
[
5] A. Yokoyama, H. Komiyama, H. Inoue, T. Masumoto, H.M.
Kimura, Scripta Metall. 15 (1981) 225.
Fig. 11. Comparison of conversion of NOx vs. time curves of the three
materials pretreated with HF.
[
[
[
6] A. Yokoyama, H. Komiyama, H. Inoue, T. Masumoto, H.M.
Kimura, J. Catal. 68 (1981) 355.
7] M. Peuckert, A. Baiker, J. Chem. Soc., Faraday Trans. 181 (1985)
streams reduces the conversion process to N2 in a few
hours, because it oxidizes the surface copper active sites.
Immersion in HF solutions produces a more effective
2
797.
8] G. Kisfaludi, K. Lazar, Z. Schay, L. Guczi, Cs. Fetzer, G. Konczos,
A. Lovas, Appl. Surf. Sci. 24 (1985) 225.
catalyst than the H treatment. This is due to the surface
[9] A. Molnar, T. Katona, M. Bartok, I .V . Perczel, Z. Hegedus, Cs.
Kopasz, Mat. Sci. Eng. A134 (1991) 1083.
10] A. Molnar, T. Katona, Cs. Kopasz, Z. Hegedus, in: L. Guczi, F.
Solymosi, P. Tetenyi (Eds.), New Frontiers in Catalysis, Elsevier
Publishers B .V , 1993, p. 1759.
2
obtained by HF treatments, which is more porous. More-
over, the Zr containing alloys show higher activity than Ti
containing materials. In fact, the oxide layers which cover
the glassy metals’ surfaces are more easily removed in the
[
[
11] T. Katona, A. Molnar, J. Catal. 153 (1995) 333.
case of Zr. Besides, the H activation of Cu–Ti amorphous
[12] T. Martinek, A. Molnar, T. Katona, M. Bartok, A. Lovas, J. Mol.
Catal. 112 (1996) 85.
2
material is ineffective in the reduction of the oxide layers.
The similar depletion observed for the catalytic activity on
the three catalysts indicates that the active sites are the
same in nature but different in number.
[13] T. Takahashi, T. Kai, Mat. Sci. Eng. A267 (1999) 207.
[14] C. Schild, A. Woukan, A. Baiker, J. Mol. Catal. 63 (1990) 243.
[15] A. Guinier, in: H.M. Foley, M.A. Ruderman (Eds.), X-Ray Diffrac-
tion in Crystals, Imperfect Crystals, and Amorphous Bodies, W.H.
Freeman and Co, San Francisco, USA, 1963, p. 61.
[16] B. Grzeta, K. Dini, N. Cowlam, H.A. Davies, J. Phys. F: Met. Phys.
1
5 (1985) 2069.
17] L. Guczi, Z. Zsoldos, G. Kisfaludi, Z. Schay, J. Mol. Catal. 51
1989) 223.
Acknowledgements
[
(
The authors would like to acknowledge the MURST for
the financial support of this work (Contract