Você está na página 1de 7

Applied Catalysis A: General 281 (2005) 199–205

www.elsevier.com/locate/apcata

Dispersion and surface states of copper catalysts by


temperature-programmed-reduction of oxidized surfaces (s-TPR)
Antonella Gervasini*, Simona Bennici
Dipartimento di Chimica Fisica ed Elettrochimica, Università degli Studi di Milano, via Camillo Golgi, 19, I-20133 Milano, Italy
Received 24 September 2004; received in revised form 22 November 2004; accepted 22 November 2004
Available online 30 December 2004

Abstract

Surface and sub-surface oxidation of dispersed copper phase by N2O adsorptive decomposition at controlled temperature followed by H2
temperature-programmed-reduction of the Cu2O surface layers formed (s-TPR) was performed on siliceous supported catalysts (ca. 6 wt.%
Cu). The combined analysis permitted to measure the copper dispersion and to identify different surface copper species. Copper dispersion
parameters were calculated from the H2-uptakes in the back-titration of the oxygen atoms fixed on the Cu particles by the s-TPR analysis. S-
shaped curves were obtained plotting the H2-uptakes versus N2O oxidation temperature, the change of slope could indicate the beginning of
copper deep oxidation, ca. 70 8C, that continued up to bulk oxidation at higher temperatures. Extrapolation of the H2-uptake to ‘‘zero-
temperature’’ allowed calculating the ‘‘true’’ copper dispersion (DCu ) and related parameters. In addition, s-TPR provided qualitative and
quantitative reduction profiles of the copper surface species. Besides Cu2O, formed by N2O oxidation of Cu(0) particles, copper species
strongly interacting with support were clearly individuated as a function of the support nature.
# 2004 Elsevier B.V. All rights reserved.

Keywords: Cu catalysts; Cu dispersion; Cu surface species; N2O oxidation; Temperature-programmed-reduction

1. Introduction it appears that a large debate is continuing from some


decades about the most suitable adsorbate to be used for the
Copper catalysts have long been used and are actually Cu dispersion determination through chemisorption meth-
used in many petrochemical processes, such as the water– ods [7–14]. The existence of weak reversible chemisorption
gas reaction and methanol synthesis. In the last decade, and uncertain adsorption stoichiometries is associated with
copper-containing zeolites and highly dispersed amorphous adsorbates, such as H2 and CO [10,12]. Low-temperature O2
copper oxide systems have been extensively studied due to chemisorption (195 or 136 8C) has been proposed as a
their exceptional activity in catalytic reactions considered as reliable technique for the measurement of Cu surface areas
potential ways for abatement of major air pollutants, e.g., [7,15], although this procedure is very time-consuming. To
NO decomposition [1–3], selective catalytic reduction of overcome this problem, O2 chemisorption at 130 8C
NOx by hydrocarbons (HC-SCR process) [4,5], and through the pulse flow technique, which should be more
decomposition of nitrous oxide [6]. suitable for routine measurements, has been adopted and
The performances of the catalysts are related to the recently described by Pernicone et al. [16]. However, O2
copper dispersion, DCu, and, therefore, the search for reliable chemisorption can be unreliable for catalysts containing low
methods for determining DCu and related parameters loadings of copper or for catalysts in which the support is
(specific Cu surface area, MSA, and average diameter of partially reducible, as pointed out in the paper of Bartley
the copper aggregates, Øav) is essential to interpret their et al. [17].
activity in terms of turnover frequencies. From the literature, The procedure, which involves adsorptive decomposition
of N2O has been used most frequently [8,9,11,14,18,19].
* Corresponding author. Tel.: +39 02 50314254; fax: +39 02 50314300. Different operating procedures have been applied to
E-mail address: antonella.gervasini@unimi.it (A. Gervasini). determine the extent of the reaction of N2O with the surface

0926-860X/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2004.11.030
200 A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205

copper: volumetric determination of the N2 produced by In this paper, an easy experimental procedure alternative
freezing out the excess of N2O [9,11], chromatographic to the others already presented in the literature [16,19,20,23]
separation of unreacted N2O and N2 formed [8,14], comprised the reliable N2O-RFC procedure [27], is
gravimetric determination of the increasing in weight of the proposed. It permits to obtain copper dispersion and specific
sample due to oxygen chemisorption [12], or a combination of copper surface area with qualitative and quantitative
methods. One of these combined procedures is based on information on the different surface copper species by
measurement of the H2 consumption during temperature- combining a surface oxidation by N2O with a successive
programmed-reduction after complete bulk oxidation and analysis of H2 temperature-programmed-reduction of the
after surface oxidation by N2O adsorption [13,18,20–22]. freshly oxidized Cu surface (defined as s-TPR, oxidized
When Cu-zeolites are concerned, copper sintering can occur surface TPR). s-TPR provides a ‘‘finger-print’’ of the surface
due to heat evolution during the N2O reaction causing copper species which can be well distinguished from the
migration and agglomeration of copper externally the zeolite bulk ones. The method has been applied to the study of
surface. In this case, temperature-programmed-desorption of conventionally prepared CuO-containing catalysts, sup-
H2 has been successfully adopted [23] to quantitatively ported on silicas of different acidity. Evidence has been
determine metallic surface areas. found that the acidity of the support strongly influences the
As regards the methodology adopted to perform the copper dispersion and copper distribution on the support.
analysis, flow as pulse methods have been adopted.
Recently, the technique termed reactive frontal chromato-
graphy (N2O-RFC) [17,24–27] has found a lot of success as 2. Experimental
particularly suitable for in situ measurements. This
technique was successfully used for the determination of Copper catalysts were prepared by impregnation of
copper areas of various catalysts used in the methanol siliceous supports with aqueous copper acetate solutions to
synthesis. In particular, a nice linear relationship between obtain samples containing about 6 wt.% Cu. The solutions
the catalytic activity of CuO/ZnO/Al2O3 catalysts and the were gently dropped on the dried powder supports. The
total copper surface areas was found [28], indicating the silicas, containing about 13 wt.% of Al2O3 (SA), TiO2 (ST),
soundness of the N2O-RFC technique. and ZrO2 (SZ), were prepared by sol–gel route [29]. All the
A problem accompanying the N2O use is the identifica- powders were dried at 120 8C for 12 h and then calcined at
tion of the correct temperature at which only the first copper 500 8C for 4 h. The Cu-samples were labelled as Cu/SA, Cu/
surface layer is oxidized without deeper Cu oxidation. ST, and Cu/SZ.
Chinchen et al. [27] have demonstrated that cutting-off the Nitrogen adsorption/desorption isotherms were measured
N2O decomposition at half-monolayer coverage, Cu bulk at 196 8C, using a Sorptometer, type Kelvin 1042, from
oxidation can be avoided. In general, temperatures between Costech International. The adsorbed volume, in
30 and 90 8C are the most frequently used to realize N2O cm3 (STP) g1, was converted into pore volume, cm3 g1,
adsorptive decomposition. However, in this temperature using the density of N2 in the normal liquid state
range, too, occurrence of some deep oxidation of copper (r = 0.8081 g cm3). The molecular area of N2 was taken
could hinder the obtainment of unambiguous results leading to be 16.2 Å2.
to overestimation of copper dispersion. Sato et al. [13] have Diffraction spectra of the powder samples without any
recently shown that slow bulk oxidation may proceed during treatment and after reduction (4.98 vol.% H2/Ar flow at
N2O oxidation even at temperature as low as 30 8C. With 15 ml min1 at 400 8C for 30 min) were collected using a
increasing N2O exposure time, a linear correlation of Cu Philips PW 1710 apparatus (Cu Ka radiation, 0.154 nm).
dispersion with t1/2 was found. At temperatures above The surface analysis was performed by XPS at room
100 8C, Cu oxidation by N2O is not restricted to the surface temperature with a M-Probe Surface Science Instrument
and bulk oxidation is quite complete [11,19]. On the other spectrophotometer with a monochromatic Al Ka radiation
hand, low dispersion values obtained for some supported Cu source operating at 1486.6 eV.
catalysts might be due to the presence of ensembles in which Copper dispersions were calculated from a two-step
the Cu atoms are not close enough to react with N2O. analysis consisting of (i) N2O oxidation of Cu to Cu2O and
Guerreiro et al. [18] suppose that a minimum distance (ii) H2 temperature-programmed-reduction of the formed
between surface Cu atoms is required in order to arrive at Cu2O surface species (s-TPR). For the analysis, a TPDRO-
N2O adsorptive decomposition, and this might lead to 1100 (from CE instruments) equipped with a quartz reactor
underestimation of metallic dispersion. The coexistence of with a porous septum (i.d., ca. 8 mm) and a filter filled with
the two opposite effects (overestimation due to deep Cu soda lime for trapping acidic gases and water was employed.
oxidation and underestimation due to not correct distance Samples ground and sieved between 45 and 60 mesh were
among Cu atom ensembles) should be ruled out to obtain used. Before the analysis, in situ pre-reduction of the CuO
reliable results. In spite of all the drawbacks discussed phase to Cu(0) was performed at 400 8C (rising the
above, the N2O decomposition is still used to measure the temperature at a rate of 8 8C min1) in a flowing H2/Ar
specific Cu surface area. mixture for 30 min (4.98 vol.% at 15 ml min1). The first Cu
A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205 201

to Cu2O oxidation by adsorptive decomposition of N2O was copper dispersion of the studied samples (DCu,A), according
carried out isothermally at different temperatures from 30 to to a method recently reported in the literature [13,18,20].
120 8C by a continuous N2O/He flow for 1 h (0.53 vol.% at
20 cm3 min1). After this, s-TPR was carried out on the
freshly oxidized Cu2O surface in order to reduce Cu2O to 3. Results and discussion
Cu. The reducing gas (4.98 vol.% H2/Ar at flow rate of
15 cm3 min1) entered into the reactor and a thermal The main characteristics of the studied samples are
conductivity detector (TCD) measured the amount of H2- presented in Table 1. The three siliceous supports onto which
uptake. Heating rates of 20 8C min1 from 40 to 800 8C copper was deposited by impregnation were high-surface
were realized. The samples, about 0.15 g, contained 120– area materials with large pore volumes [29,32]. The surfaces
140 mmol of Cu to maintain K- and P-values [30,31] around possessed variable concentration of acid sites which nature,
150 s and 20 8C, respectively. Quantitative H2-uptakes were and acid strength depended on co-oxide in the silica
evaluated by integration of the experimental s-TPR curves, structure. As expected, silica–alumina (SA) and silica–
basing on previous calibration measurements with CuO zirconia (SZ) had higher acidity than silica–titania (ST),
powder (Fluka, >98 wt.% purity). both in terms of amount of acid sites and mean acid strength
Copper dispersion (DCu), defined as the ratio of Cu of their surfaces [29].
exposed at the surface to total Cu, was calculated from the The impregnated catalysts had all similar Cu content (ca.
amount of H2 consumed in the s-TPR analysis. Starting from 6 wt.%) and high-surface area (300–500 m2 g1). In all the
the DCu value, Cu metal surface area, MSA, and average cases, the calcined catalysts presented a XRD pattern in
copper particle size, Øav were calculated. Metal surface area which, besides a very broad band at 20–308 2u, typical of
was calculated as: MSA (m2 g1 Cu ) = MolH2 SFNA/ unstructured silica groups, only two definite peaks at 358 and
104CMWCu, where MolH2 , SF, NA, CM, and WCu are moles 388 2u appeared. The two peaks are typical of the CuO
of hydrogen experimentally consumed per unit mass of presence. On the other hand, the reduced catalysts showed
catalyst (mmolH2 g1 cat ), stoichiometric factor (2), Avogadro’s more definite X-ray peaks at 438 and 508 2u, typical of the
number (6.022  1023 mol1), number of surface Cu atoms Cu(0) phase. Fig. 1 displays the XRD spectrum of Cu/SZ in
per unit surface area (1.47  1019 atoms m2), and Cu the reduced and unreduced states, as an example. A first
content (wt.%), respectively. The O/Cu ratio is assumed to approximate estimate of the CuO particle dimensions was
be 1/2 (SF = 2) on the basis of UPS results [14], which performed from the broadening of the peak at 358 2u by the
proved that after oxidation with N2O at temperatures up to Scherrer law. CuO dimensions of about 15 nm could be
100–120 8C, the surface copper is primarily in the CuI calculated, in any case. The same procedure was applied to
oxidation state. The CM value of 1.47  1019 atoms m2, the 438 2u peak of the reduced copper sample, obtaining
supposing equal areas of exposed (1 0 0), (1 1 0), and (1 1 1) dimensions in the 20–30 nm interval for the three samples.
planes, i.e., 0.065, 0.092, and 0.0563 nm2 [14,18], results in XP spectra revealed surfaces enriched with copper;
an average Cu surface area of 0.0711 nm2 per atom. Average signals at binding energy values around 934 eV (Cu 2p3/2)
copper particle size was calculated as: Øav with very intense and broad satellite peaks at 943 eV were
(nm) = 107SKCMWCu/SFMolH2 NArCu, where SK is a observed on all the surfaces. From these findings, it can be
constant depending on Cu particle shape (6 or 5) and rCu is deduced that in the fresh catalysts there is the sole Cu(II)
the density of copper (8.92 g cm3) [18]. presence [33,34]. Two contributions could be recognized in
Conventional H2 temperature-programmed-reduction the 934 eV convoluted band, the one centered at 933 and the
analysis (TPR) was performed under the same experimental second at 936 eV. In agreement with similar observations
conditions (20 8C min1 from 40 to 800 8C) above reported reported by Córdoba et al. [35] and Bennici et al. [33], the
on the fully oxidized CuO samples obtained by air treatment two signals can be associated to CuO (933 eV) and to Cu–O–
(flow rate of 45 cm3 min1) at 350 8C for 1 h. The ratio Si–O– (936 eV) species. For the catalyst prepared on the
between the s-TPR (obtained after N2O oxidation at 30 8C) silica–alumina support (Cu/SA), the presence of Cu(II) ions
and TPR peak areas directly furnished another evaluation of in aluminate-like compounds cannot be ruled out. In fact,

Table 1
Bulk and surface characteristics of the dispersed copper catalysts
Code Cu content SA (m2 g1) DCu (%)a MSA* (m2 g1
Cu )
a ? av (nm)a,b
(wt.%) (mmol g1
cat )
Cu/SA 5.3 0.834 481 4.8 30.9 21.8
Cu/ST 6.1 0.960 340 4.7 30.6 22.0
Cu/SZ 5.9 0.928 278 5.8 37.5 17.9
a
Determined from H2-uptake extrapolated to ‘‘zero-temperature’’ (see Section 2).
b
Calculated with SK = 6 (spherical or cubic shape of the Cu aggregates).
202 A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205

Fig. 2. Illustration of the stepwise procedure used.


Fig. 1. X-ray diffraction patterns of Cu/SZ: (a) fresh (CuO) and (b) reduced
(Cu(0)) samples.
quantified. As described in Fig. 2, the concentration profile
of N2O, flowing through the powder sample, lowered
recent results of Shimizu et al. [36] indicate that Cu2+ ions in following N2O adsorption and then it returned to its initial
CuAl2O4 exhibit an XPS peak around 935 eV. value once all the exposed copper sites were oxidized to
Central experiments performed on the studied catalysts Cu2O. The N2O consumed was indirectly evaluated by the
were aimed at quantitatively determining the copper surface H2-back-titration (step 2 analysis, s-TPR) of the oxygen
dispersion degree. In Table 2 and Fig. 2, we described the atoms left on the surface copper particles (Fig. 2).
two-step procedure followed for the determination of copper The s-TPR profiles (step 2) obtained after N2O oxidation
dispersion. Before the measurements, an appropriate performed at variable temperatures are reported in Fig. 3b.
experimental procedure was applied to the catalysts in The profiles show two well distinct peaks at low (150–
order to reduce the CuO phase to metallic phase (step 0, see 200 8C) and higher temperatures (500–600 8C), over all
Table 2). The choice of the experimental conditions for step three catalysts. The low-temperature peaks did not
0 obviously depends on the type of sample; the conditions remarkably differ in shape and intensity when registered
have to be optimized in each case. In the present case, the after N2O oxidation at temperatures in the 30–50 8C interval,
pre-reducing treatment guaranteed the absence of copper while, for higher oxidation temperatures (up to 120 8C) a
sintering, at the utmost, a very modest copper aggregation very definite increasing of peak intensity was observed. The
could occur. In fact, the copper particle size of the reduced high-temperature peaks were well pronounced only for Cu/
samples, calculated from the relevant XRD spectra, was SA, with maxima around 540 8C. The peak areas were not
similar to that of the calcined CuO samples, above sensitive to the temperature at which the N2O oxidation was
described. performed in the overall interval explored (30–120 8C). The
Step 1 of the combined procedure led to surface oxidation different behaviour of the low- and high-temperature s-TPR
or to surface and sub-surface oxidation of the Cu particles, peaks in relation to N2O oxidation temperature suggests that
depending on the temperature chosen [11,14,19]. N2O the low-temperature peak is due to the reduction of Cu2O
exposure time also influences the extent of sub-surface species. As higher the oxidation temperature was as more
oxidation, all the more so on choosing a higher oxidation intense the peak was (higher H2-uptake was observed),
temperature [13]. In the present work, the N2O exposure because sub-layer copper oxidation could occur. The high-
time was maintained constant (1 h) and only the temperature temperature peak could be due to the reduction of copper
was varied. The consumption of N2O was not directly compounds, probably formed during catalyst preparation
and involving chemical species of the support (e.g.,
CuAl2O4-like compounds). Such Cu-containing compounds
Table 2 were difficult to be reduced, as indicates the high-
Stepwise procedure of the combined technique temperature value of the reduction peak. Taking into
Step Procedure account the total amount of copper in the Cu/SA sample,
0 Pre-reducing treatment with H2/Ar (4.98 vol.%, 15 ml min1) it can be deduced that about 10% of copper was bound to the
at 400 8C (8 8C min1) for 30 min, cooling to 40 8C. SA support and it was not available as free copper. The mild
Supported copper particles, Cu(0), are formed conditions used for step 0 were effective for reducing the
1 Isothermal surface oxidation by N2O/He (0.53 vol.%, CuO phase to Cu(0) but they did not allow reducing the
20 ml min1) in the 30–120 8C range for 1 h formed CuAl2O4-like compounds, as emerged from the
Surface oxygen besides sub-surface oxygen is taken up reduction profiles of the catalysts collected after pre-
(Cu2O formation) reducing treatment (Fig. 3a). In particular for Cu/SA, a large
2 Temperature-programmed-reduction (s-TPR) in H2/Ar and intense peak was still present (at ca. 540 8C) after the
(4.98 vol.%, 15 ml min1) from 40 to 800 8C at 20 8C min1 pre-reducing treatment. The area of this peak corresponded
H2-uptake is determined (Cu2O ! Cu(0) reduction)
to an H2 consumption of about 90 mmol g1, which is very
A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205 203

oxidation. Fig. 4 shows the results of H2-back-titration of the


oxygen atoms fixed on the Cu particles by s-TPR. The
obtained curves (H2-uptakes versus N2O oxidation tem-
perature) were S-shaped for the all the three samples. The
hydrogen uptake very slightly increased by increasing the
temperature of oxidation from 30 to 50 8C or to 70 8C for the
Cu/SZ and Cu/SA samples, respectively. Such behaviour
suggests that in this temperature range only the surface
layers of Cu2O are involved in the reduction. Above 50 8C,
for Cu/SZ, or above 70 8C for Cu/SA, there was a sharp
increasing in the hydrogen uptake indicating the reduction of
sub-surface layers. On the other hand, the Cu/ST sample did
not show any constant range of H2-uptake against N2O
oxidation temperature. The peculiar shape of the curves of
Fig. 4, can suggest that the change of slope of the S-shaped
curves indicates the onset of copper bulk oxidation.
The copper dispersion was calculated from the H2-uptakes
determined from the s-TPR analyses, as above described (see
Section 2). For Cu/SA, DCu values around 5% were obtained
following temperatures of Cu oxidation by N2O in the range
32–76 8C; whereas, a value of 17% was calculated following
temperatures of N2O oxidation of 108 8C. Analogously for Cu/
SZ, a quite unique value of DCu (6%) was observed following
temperatures of copper oxidation by N2O in the 29–48 8C
interval, while higher values, 10–12%, were obtained
following oxidation temperatures in the 80–120 8C interval.
For Cu/ST, DCu showed strongly increasing values (from 5 to
14%) in all the temperature range of N2O oxidation
investigated (35–120 8C). Due to the ability of N2O to oxidize
the copper surface and the sub-surface layers depending on the
oxidation temperature chosen, the problem of overestimation
of the copper dispersion arises. To limit this problem,
extrapolation to ‘‘zero-temperature’’ of the curves of Fig. 4,
could allow to obtain true copper dispersion (DCu ). The
procedure enables to avoid overestimation of the Cu
dispersion, even if in some cases the Y-intercept of such
curves could be questionable. The DCu values with the relevant
specific copper surface areas, MSA*, and average dimensions
Fig. 3. Reduction profiles of Cu/SA (A), Cu/ST (B), and Cu/SZ (C): (a) of copper aggregates. ? av , calculated from the hydrogen-
TPR after the pre-reducing treatment (step 0 analysis); (b) s-TPR of freshly
oxidized Cu2O phase (step 2 analysis) after N2O oxidation at various
uptakes extrapolated to ‘‘zero-temperature’’ are reported in
temperatures (step 1 analysis). Table 1. The DCu values, that reflect free copper surfaces, were
in the order Cu/ST  Cu/SA < Cu/SZ. Moreover, the average
copper particle size of the three samples (? av = 18–22 nm) is
similar to that of the high temperature s-TPR peak. Over Cu/ well within the same interval of size obtained from the XRD
ST and Cu/SZ, the high-temperature peaks are of very low analysis (20–30 nm) for the reduced catalysts.
intensity, indicating very small formation of Cu-containing The studied catalysts have copper dispersion similar to
compounds during the preparation step. It is hard to know if other copper catalysts reported in the literature [13,16,24]
the high-temperature peak is associated with exclusive and not important differences emerge among the studied
presence of surface copper phase. samples when one considers the free copper phase (Cu/SZ
In the light of the evidences emerged from the s-TPR shows higher value of about 20% than Cu/SA and Cu/ST).
profiles, we decided to consider only the s-TPR peaks at low Thus, the preparation method seems to govern the copper
temperatures, related with Cu2O to Cu(0) reduction, in order dispersion rather than the support nature. As concerns the
to determine the values of copper dispersion (DCu) of all the influence of the support nature, the high acidity of SA, which
samples. is due to Al presence, caused an appreciable copper
As Fig. 3b shows, increasing Cu2O peak areas were clustering (about 10%) due to formation of Cu-containing
obtained with the increasing of the temperature of N2O stable compounds.
204 A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205

Fig. 4. Influence of the temperature of N2O oxidation reaction (step 1 analysis) on hydrogen-uptake (step 2 analysis) for Cu/SA (A), Cu/ST (B), and Cu/SZ (C).

However, a different picture can be obtained if one For comparative purposes and to corroborate the obtained
considers the integration of both the low- and high- results, the copper dispersion has been calculated according
temperature peaks of the s-TPR profiles to calculate copper to a procedure used in the literature based on the
dispersion of the three catalysts (Fig. 5). The new DCu values measurement of the H2 consumption after complete
obtained after N2O oxidation at 30 8C of the copper surfaces oxidation of the catalyst and after surface oxidation by
and calculated attributing the area of the high-temperature N2O [13,18,20]. By applying this method to the present case,
peak to surface Cu(II) species, are reported in Fig. 5. The the ratio between peak areas of conventional TPR profile
contribution of the high-temperature peak dramatically (A1), corresponding to total amount of Cu(II) in the sample,
influences the copper dispersion value of Cu/SA. A lighter and that of s-TPR (A2) (low-temperature s-TPR peak only),
increasing of Cu dispersion could be calculated for Cu/SZ and corresponding to Cu(I) phase formed by N2O oxidation,
Cu/ST. This is a clear evidence of the influence of the support directly can give the value of copper dispersion,
nature on the copper aggregation state. Preferential clustering DCu,A = 2A2/A1. The values observed for DCu,A and DCu,
of copper on alumina rich regions is likely to occur with when compared after N2O oxidation of the copper phase at
interaction between copper and alumina leading to Cu species 30 8C, are lightly different. DCu,A values (5.8% for Cu/SA,
more difficult to be reduced than free CuO species. 5.6% for Cu/ST, and 6.8% for Cu/SZ) are all higher than DCu
A. Gervasini, S. Bennici / Applied Catalysis A: General 281 (2005) 199–205 205

Ragaini for helpful suggestions during the course of this


work.

References

[1] M. Iwamoto, H. Hamada, Catal. Today 10 (1991) 57.


[2] M. Iwamoto, H. Yahiro, N. Mizuno, S. Kagawa, Chem. Lett. (1989)
213.
[3] A. Gervasini, Appl. Catal. B 14 (1997) 147.
[4] M. Shelef, Chem. Rev. 95 (1995) 209.
[5] J.A. Anderson, C. Márquez-Alvarez, M.J. López-Munoz, I. Rodri-
guez-Ramos, A. Guerrero-Ruiz, Appl. Catal. B 14 (1997) 189.
Fig. 5. Comparison between the H2-uptakes calculated by integration of the
[6] A. Dandekar, M.A. Vannice, Appl. Catal. B 22 (1999) 179.
low temperature and the low- and high-temperature s-TPR peaks after N2O
[7] A.A. Vasilevich, G.P. Shpiro, A.M. Alekseev, T.A. Semenova, M.I.
oxidation at 30 8C. The relevant percent copper dispersion values are
Markina, T.A. Vasil’eva, O.G. Budkina, Kinet. Catal. 16 (1975) 1363.
indicated.
[8] B. Dvorak, J. Pasek, J. Catal. 18 (1970) 108.
[9] G. Sengupta, D.K. Gupta, M.L. Kundi, S.P. Sen, J. Catal. 67 (1981)
223.
(4.8% for Cu/SA, 5.3% for Cu/ST, and 5.8% for Cu/SZ). [10] M. Muhler, L.P. Nielsen, E. Tornqvist, B.S. Clausen, H. Topsoe, Catal.
Non-quantitative reduction of CuO to Cu(0) during Lett. 14 (1992) 241.
conventional TPR analysis can be responsible of this [11] Th.J. Osinga, B.G. Linsen, W.P. van Beek, J. Catal. 7 (1967) 277.
behaviour, leading to underestimation of the A1 peak areas. [12] A. Dandenkar, M.A. Vannice, J. Catal. 178 (1998) 621.
[13] S. Sato, R. Takahashi, T. Sodesawa, K. Yuma, Y. Obata, J. Catal. 196
(2000) 195.
[14] J.W. Evans, M.S. Wainwright, A.J. Bridgewater, D.J. Young, Appl.
4. Conclusion Catal. 7 (1983) 75.
[15] G.E. Parris, K. Klier, J. Catal. 97 (1986) 374.
Oxidized surface layers of copper phase (Cu2O) was [16] N. Pernicone, T. Fantinel, C. Baldan, P. Riello, F. Pinna, Appl. Catal. A
240 (2003) 199.
reduced (s-TPR) allowing the measurement of specific Cu [17] G.J.J. Bartley, R. Burch, R.J. Chappell, Appl. Catal. 43 (1988)
surface area and identification of different surface copper 91.
species. The used method was applied to catalysts prepared [18] E.D. Guerreiro, O.F. Gorriz, J.B. Rivarola, L.A. Arrúa, Appl. Catal. A
by conventional impregnation, which is still a convention- 165 (1997) 259.
[19] J.J.F. Scholten, J.A. Konvalinka, Trans. Faraday Soc. 65 (1969) 2465.
ally adopted preparation in chemical industry. The method
[20] G.C. Bond, S.N. Namijo, J. Catal. 118 (1989) 507.
consists in an isothermal surface oxidation by N2O at [21] Y.-J. Tu, Y.-W. Chen, Ind. Eng. Chem. Res. 37 (1998) 2618.
definite temperature followed by H2-back-titration (s-TPR), [22] E.D. Guerreiro, O.F. Gorriz, G.L. Larsen, L.A. Arrúa, Appl. Catal. A
it can be suitable for routine measurements of copper 204 (2000) 33.1.
dispersion. Homogeneous series of copper catalysts may be [23] M.C.N. Amorim de Carvalho, F.B. Passos, M. Schmal, Catal. Comm. 3
comparatively studied by performing the N2O surface (2002) 503.
[24] M.M. Günter, T. Ressler, B. Rems, C. Büscher, T. Genger, O.
oxidation in low-temperature range (30–50 8C), and then Hinrichsen, M. Muhler, R. Schlögl, Catal. Lett. 71 (2001) 37.
counting the oxygen left at the surface. Due to the high [25] R.N. d’Alnoncourt, M. Kurtz, H. Wilmer, E. Löffler, V. Hagen, J. Shen,
sensitivity of the two-step technique, catalysts at low Cu M. Muhler, J. Catal. 220 (2003) 249.
loading, as those used for environmental purposes, could be [26] J. Agrell, M. Boutonnet, I. Melián-Cabrera, J.L.G. Fierro, Appl. Catal.
studied. A 253 (2003) 201.
[27] G.C. Chinchen, C.M. Hay, H.D. Vandervell, K.C. Waugh, J. Catal. 103
Moreover, the qualitative and quantitative determination (1987) 79.
of different surface copper species can be performed by s- [28] G.C. Chinchen, K.C. Waugh, D.A. Whan, Appl. Catal. 25 (1986) 101.
TPR analysis. Knowledge of the type and amount of the [29] P. Carniti, A. Gervasini, S. Bennici, J. Phys. Chem. B (2005) in press.
surface copper species may be of some interest for the [30] P. Malet, A. Caballero, J. Chem. Soc., Faraday Trans. 1 84 (1988)
2369.
calculation of the turnover-frequency values when inter-
[31] D.A.M. Monti, A. Baiker, J. Catal. 83 (1983) 323.
pretation of catalytic performances for reactions catalysed [32] S. Bennici, P. Carniti, A. Gervasini, Catal. Lett. 98 (2004) 187.
by copper phases is desired. [33] S. Bennici, A. Gervasini, N. Ravasio, F. Zaccheria, J. Phys. Chem. B
107 (2003) 5168.
[34] J. Morales, A. Caballero, J.P. Holgado, J.P. Espinós, A.R. González-
Acknowledgements Elipe, J. Phys. Chem. B 106 (2002) 10185.
[35] G. Córdoba, M. Viniegra, J.L.G. Fierro, J. Padilla, R. Arroyo, J. Solid
State Chem. 138 (1998) 1.
The authors wish to thank Ms. Iolanda Biraghi for [36] K. Shimizu, H. Maeshima, H. Yoshida, A. Satsuma, T. Hattori, Phys.
performing XRD measurements and Professor Vittorio Chem. Chem. Phys. 2 (2000) 2435.

Você também pode gostar