effect of pretreatment conditions on the catalytic performance of ni–pt–w supported on amorphous...

9
Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts Part 1. Catalysts prepared by a sol–gel method Yacine Rezgui * , Miloud Guemini Laboratoire de Recherche de Chimie Applique ´e et Technologie des Mate ´riaux, Universite ´ d’Oum El Bouaghi, B.P. 358, Route de Constantine, Oum El Bouaghi 04000, Algeria Received 19 September 2007; received in revised form 10 November 2007; accepted 14 November 2007 Available online 22 November 2007 Abstract In this work, the catalytic behavior of Pt–Ni–WO x supported on amorphous silica–alumina catalysts, prepared by means of a sol–gel technique was tested by hydroisomerization of n-hexane in a continuous fixed-bed reactor operating at atmospheric pressure. Temperature-programmed desorption of ammonia (TPDA), temperature-programmed reduction, BET and atomic absorption spectroscopy techniques were used to characterize the catalysts. The results revealed that both catalytic activity and surface acidity of these solids is strongly dependent on their calcination and reduction temperatures. Besides, the collected data showed that platinum interacts with tungsten and this interaction depends on the pretreatment conditions. # 2007 Elsevier B.V. All rights reserved. Keywords: Platinum; Tungsten oxide; Nickel; Isomerization; Reduction; Acidity 1. Introduction Environmental and public health reasons are motivating the production of transportation fuels with increase amount of high octane number branched alkanes [1]. These compounds are made by alkylation and by isomerization of straight-chain alkanes. The drawback of these two processes is that they employ harmful and potentially polluting catalysts, such as H 2 SO 4 , and HF [2]. Nowadays, most working isomerization plants use Pt/Cl–Al 2 O 3 catalysts, that have good activity and selectivity at relatively low temperatures (T reac = 110–135 8C) [3]. These catalysts are however candidates for replacement because of their corrosiveness and potential for pollution. Active, stable solid acid catalysts that might replace chlorinated alumina are the halid-free solid acid catalysts. Among them, the best candidate was sulfated zirconia, this latter is actually used in industry [4], but it is said that it may undergo sulfur loss during reaction [5] leading to structural modifications. There- fore, the amelioration or the replacement of this solid by some other superacid catalyst of higher stability, especially for hydrocarbons heavier than butane, is currently desired. Recently bulk tungsten oxide-based catalysts have attracted considerable attention, these solids seems to be potential catalysts for skeletal rearrangements of hydrocarbons [6–10]. The presence of noble metals (Pt or Pd) in such solids was expected to improve their stability but also to activate the isomerization toward saturated hydrocarbons [11–13]. How- ever it was mentioned that, when introducing noble metals directly to bulk tungsten oxides, the metallic function was very sensitive to hydrocarbon poisoning, leading to catalyst deactivation toward saturated alkanes. Thus the necessity to stabilize the bulk oxide and metallic active phases was mentioned and the use of a support seemed to be a potential way [10]. Based on these facts, it would be said that tungsten catalytic performance can be largely modified upon dispersing on oxide supports, where variables such as calcination temperature, nature of the supported phase and surface acidity–basicity as well as the nature of the support used are of paramount importance [14,15]. As mentioned by Maciej and Zbigniew [16], catalytic behavior of supported metals in the conversion of saturated hydrocarbons can be influenced by components, metal and support. The relative contribution of the www.elsevier.com/locate/apcata Available online at www.sciencedirect.com Applied Catalysis A: General 335 (2008) 103–111 * Corresponding author. E-mail address: [email protected] (Y. Rezgui). 0926-860X/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2007.11.020

Upload: yacine-rezgui

Post on 26-Jun-2016

221 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

www.elsevier.com/locate/apcata

Available online at www.sciencedirect.com

Applied Catalysis A: General 335 (2008) 103–111

Effect of pretreatment conditions on the catalytic performance

of Ni–Pt–W supported on amorphous silica–alumina catalysts

Part 1. Catalysts prepared by a sol–gel method

Yacine Rezgui *, Miloud Guemini

Laboratoire de Recherche de Chimie Appliquee et Technologie des Materiaux, Universite d’Oum El Bouaghi, B.P. 358,

Route de Constantine, Oum El Bouaghi 04000, Algeria

Received 19 September 2007; received in revised form 10 November 2007; accepted 14 November 2007

Available online 22 November 2007

Abstract

In this work, the catalytic behavior of Pt–Ni–WOx supported on amorphous silica–alumina catalysts, prepared by means of a sol–gel technique

was tested by hydroisomerization of n-hexane in a continuous fixed-bed reactor operating at atmospheric pressure. Temperature-programmed

desorption of ammonia (TPDA), temperature-programmed reduction, BET and atomic absorption spectroscopy techniques were used to

characterize the catalysts. The results revealed that both catalytic activity and surface acidity of these solids is strongly dependent on their

calcination and reduction temperatures. Besides, the collected data showed that platinum interacts with tungsten and this interaction depends on the

pretreatment conditions.

# 2007 Elsevier B.V. All rights reserved.

Keywords: Platinum; Tungsten oxide; Nickel; Isomerization; Reduction; Acidity

1. Introduction

Environmental and public health reasons are motivating the

production of transportation fuels with increase amount of high

octane number branched alkanes [1]. These compounds are

made by alkylation and by isomerization of straight-chain

alkanes. The drawback of these two processes is that they

employ harmful and potentially polluting catalysts, such as

H2SO4, and HF [2]. Nowadays, most working isomerization

plants use Pt/Cl–Al2O3 catalysts, that have good activity and

selectivity at relatively low temperatures (Treac = 110–135 8C)

[3]. These catalysts are however candidates for replacement

because of their corrosiveness and potential for pollution.

Active, stable solid acid catalysts that might replace chlorinated

alumina are the halid-free solid acid catalysts. Among them, the

best candidate was sulfated zirconia, this latter is actually used

in industry [4], but it is said that it may undergo sulfur loss

during reaction [5] leading to structural modifications. There-

fore, the amelioration or the replacement of this solid by some

* Corresponding author.

E-mail address: [email protected] (Y. Rezgui).

0926-860X/$ – see front matter # 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.apcata.2007.11.020

other superacid catalyst of higher stability, especially for

hydrocarbons heavier than butane, is currently desired.

Recently bulk tungsten oxide-based catalysts have attracted

considerable attention, these solids seems to be potential

catalysts for skeletal rearrangements of hydrocarbons [6–10].

The presence of noble metals (Pt or Pd) in such solids was

expected to improve their stability but also to activate the

isomerization toward saturated hydrocarbons [11–13]. How-

ever it was mentioned that, when introducing noble metals

directly to bulk tungsten oxides, the metallic function was very

sensitive to hydrocarbon poisoning, leading to catalyst

deactivation toward saturated alkanes. Thus the necessity to

stabilize the bulk oxide and metallic active phases was

mentioned and the use of a support seemed to be a potential way

[10]. Based on these facts, it would be said that tungsten

catalytic performance can be largely modified upon dispersing

on oxide supports, where variables such as calcination

temperature, nature of the supported phase and surface

acidity–basicity as well as the nature of the support used are

of paramount importance [14,15]. As mentioned by Maciej and

Zbigniew [16], catalytic behavior of supported metals in the

conversion of saturated hydrocarbons can be influenced by

components, metal and support. The relative contribution of the

Page 2: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Table 1

Chemical composition of the (NixPtyW/ASA)BC samples (noted (NixPty)BC)

Catalyst %Ni %Pt %W SiO2/Al2O3

Ni12Pt0 12 0.0 10 1.83

(Ni12Pt0.1)BC 12 0.1 10 1.83

(Ni12Pt0.4)BC 12 0.4 10 1.83

(Ni12Pt1)BC 12 1.0 10 1.83

Ni15Pt0 15 0.0 10 1.83

(Ni15Pt0.1)BC 15 0.1 10 1.83

(Ni15Pt0.4)BC 15 0.4 10 1.83

(Ni15Pt1)BC 15 1.0 10 1.83

Ni17Pt0 17 0.0 10 1.83

(Ni17Pt0.1)BC 17 0.1 10 1.83

(Ni17Pt0.4)BC 17 0.4 10 1.83

(Ni17Pt1)BC 17 1.0 10 1.83

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111104

support to the overall reactivity is generally attributed to its

surface acidity which could be enhanced by thermal treatments.

These thermal pretreatments are related to the catalyst

preparation method. Several methods for the preparation of

supported metal oxide catalysts have been reported in the

literature: mechanical mixing of metal compound and support,

ion exchange processes, coprecipitation of metal oxides and

support, impregnation and incipient wetness processes [17–21],

solid–solid wetting [22] and grafting of metal alkoxide

precursors onto a metal oxide support alkoxides [23–26].

A great deal of attention in the last two decades has been

given to the use of sol–gel method which is a homogeneous

process resulting in a continuous transformation of a solution

into hydrated solid precursor (hydrogel). The gelation route

(sol–gel process) has several promising advantages over the

conventional techniques. It offers better control of the texture,

composition, homogeneity and structural properties of the final

solids [27–29].

In a previous paper, we have used the latter method to

synthesize catalysts based on nickel–tungsten oxide supported

on silica–alumina [30] and we have demonstrated that these

materials can exhibit an interesting behavior in catalytic

hydroisomerization (up to 79.2% and 70% selectivity to

isomers in the case of n-hexane and n-heptane, at 24% and 29%

conversions, respectively) [30,31] and dewaxing (the selectiv-

ity to isomers, in the case of n-decane, amounts to 55% at a

conversion of 42.3%) [32] if the rate of hydrogenolysis can be

diminished. In this aim, the objective of our work was the

incorporation, by different methods, of platinum in Ni–WOx/

SiO2–Al2O3 catalysts and the attempt to study the effect of

pretreatment conditions on the activity of the prepared solids.

This issue seems important in light of common practice of using

alkanes (n-hexane) as test molecules for probing the state of

metal in supported catalysts. In this first part, we decided to

study the catalytic behavior of (NixPty)BC samples, prepared by

sol–gel method where the platinum was incorporated before

calcination.

2. Experimental

2.1. Catalyst preparation

We studied the effects of the pretreatment conditions on the

activity of Ni–Pt–WOx/SiO2–Al2O3 catalysts for n-hexane

hydroisomerization. In this order, nine catalysts were prepared

and were treated with different conditions. The prepared solids

noted (NixPty)BC (Table 1), where x and y indicate the

percentage of nickel and platinum in the catalyst, and BC

indicate that the incorporation of Pt was done before

calcination, having the same amount of tungsten (10%) and

a constant SiO2/Al2O3 ratio (1.83), were prepared using a sol–

gel method.

A sol was obtained by mixing, under vigorous stirring, an

aqueous solution of nickel nitrate (Ni(NO3)2�6H2O) prelimi-

narily acidified by nitric acid with the required amounts of

aluminum sulfate (Al2(SO4)3�18H2O), sodium tungstate

(Na2WO4) and hexachloroplatinic acid (H2PtCl6) solutions.

To the sol obtained, under vigorous stirring, an aqueous

solution of sodium silicate (Na2SiO3) was added. Next,

ammonium sulfate was used to activate the prepared gel at

60 8C (liquid to solid ratio equal to 30), under reflux conditions

in a thermostat and over a period of 48 h (this unit operation

was repeated several times). Then the solid was washed with

hot water (60 8C), dried at 120 8C for 4 h and finally calcined in

the temperature range 300–700 8C for 5 h. A heating rate of

10 8C/min was used.

For comparison purposes, Pt-free catalysts containing 12%,

15%, 17% Ni and 10% of tungsten supported on amorphous

silica–alumina were prepared as described in Ref. [31].

2.2. Catalyst characterization

The specific surface area of the catalysts was determined by

the BET method. The adsorption isotherm was measured at the

temperature of liquid nitrogen. Before the adsorption was

measured, all samples were degassed at 350 8C for 30 min, and

then reduced at 430 8C, with hydrogen, for 1 h. The tungsten,

nickel and platinum concentration was measured by inductively

coupled plasma-atomic emission spectroscopy (ICP-AES).

The surface species reducibility was determined by

temperature-programmed reduction (TPR) using an Ohkura

TP 2002 S equipped with a thermal conductivity detector. After

loading, samples were pretreated at 500 8C in an air flow for 3 h

and afterward cooled at room temperature. Next, they were

stabilized in an Ar/H2 (95/5 volumetric ratio) flow. The

temperature and detector signals were then continuously

recorded while heating at 5 8C/min to 600 8C.

The acidic properties of samples were studied by means of

temperature-programmed desorption of ammonia (TPD of

NH3). Prior to TPD experiments, the sample was pretreated, for

3 h at the desired calcination temperature, with an oxygen flow,

then purged by flowing helium at 300 8C, and then reduced, at

430 8C for 2 h, in a hydrogen flow. After reduction, the sample

was further dried in flowing He at 300 8C for 1 h and then

cooled to room temperature. When the system become steady,

ammonia was adsorbed at 100 8C for 30 min (using a 10% NH3/

He carrier gas) and then the sample was subsequently purged, at

the same temperature, by flowing He (100 ml/min) for 1 h to

remove the excess and physically adsorbed NH3. The TPD

Page 3: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111 105

spectrum was obtained by heating the sample from 100 to

700 8C at a heating rate of 10 8C/min under a He flow. Evolved

NH3 was monitored with a TCD detector.

2.3. Catalytic test

The n-hexane transformation was performed in a continuous

flow fixed-bed quartz reactor, loaded with 1 g of the catalyst,

operated under isothermal conditions, and heated by a

controlled temperature electrical oven. The reaction was

carried out at atmospheric pressure, 250 8C, weight hourly

space velocity (WHSV) = 4 h�1 and a molar ratio of hydrogen/

hydrocarbon (H2/n-hexane) = 5.

The reaction products were analyzed on-line by gas

chromatography.

Fig. 1. Effect of calcination temperature on the performance of (a) (Ni12Pty)BC, (b) (

on stream (TOS) = 100 min.

3. Results and discussion

3.1. Effect of calcination temperature

The effect of calcination temperature on the catalytic

activities of the prepared solids was investigated at a reaction

temperature of 250 8C, for a time on stream of 100 min and at

various reduction temperatures (Tred). As shown in Figs. 1 and

2a and c, whatever the reduction temperature, the Ni and Pt

amounts, the catalytic activity of all the prepared samples

increased with calcination, passed through a maximum and

then decreased. Calcination at 700 8C yielded the lowest

activity. The catalyst activities depend strongly on the

calcination temperature, reduction temperature as well as on

the nickel and platinum concentrations. The calcination

Ni15Pty)BC, and (c) (Ni17Pty)BC catalysts. Tred = 260 8C, Treac = 250 8C and time

Page 4: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Fig. 2. Effect of calcination temperature on the performance of (a) (Ni12Pty)BC, (b) (Ni15Pty)BC, and (c) (Ni17Pty)BC catalysts. Tred = 440 8C, Treac = 250 8C and time

on stream (TOS) = 100 min.

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111106

temperatures showing the maximum activity were higher with

less nickel and platinum, which suggest that interaction of

nickel and platinum with the support surface were strong when

the active species are present in small amounts, similar trends

were reported by Hino and Arata [33] in their study on the

isomerization of n-butane over (WO3/ZrO2 + Pt/ZrO2) cata-

lysts. Besides, regardless the reduction temperature and the

nickel amount, an increase in Pt content induces a small

increase in the maximum conversion of the prepared samples.

Fig. 3a gives the dependence of the catalysts acidities on

their calcination temperatures. All samples showed virtually

identical trends with increasing calcination temperature in that

their acidities increased to reach a maximum and then they

decreased. If we combine the data collected in Fig. 2 and these

results, we can say that the catalytic performance of the

prepared materials is strongly dependent on their acid sites

density, the higher the acid sites concentration the higher the

activity showed by the catalyst. On the other hand, the collected

data showed that the BET surface areas of the prepared

catalysts were not influenced by their calcination temperatures

(Fig. 3b).

Hereafter and for labeling purposes, we specify the wt.% of

either Ni or Pt immediately after the chemical symbol as

subscripts, and the optimum calcination temperature (8C) (see

Table 2) as the label suffix. For example, the 12 wt.% Ni and

0.1 wt.% Pt catalyst that had been calcined at 500 8C was

labeled as (Ni12Pt0.1)BC-500.

3.2. Reduction behavior

Fig. 4a–c displays TPR patterns obtained for (NixPty)BC

catalysts. As it clearly can be seen, these solids exhibit an

Page 5: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Fig. 3. (a) Dependence of the catalysts acidities on their calcination temperatures at Tred = 430 8C. (b) Dependence of the catalysts BET surface areas on their

calcination temperatures at Tred = 430 8C.

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111 107

increase in hydrogen consumption in the temperature range

276–371 8C, followed by two peaks centered around 420, 400,

385 and 405 8C (for the first peak) and 460, 471, 477 and

465 8C (for the second peak), for Ni12Pt0, (Ni12Pt0.1)BC,

(Ni12Pt0.4)BC and (Ni12Pt1)BC catalysts, respectively. On the

other hand, the Ni15Pt0 catalyst consumes hydrogen in the

temperature range 265–375 8C and it shows a peak at about

410 8C followed by a shoulder at about 440 8C, whereas the

other (Ni15Ptx)BC (x = 0.1, 0.4 and 1) display an increase in the

base line in the temperature range 253–372 8C, followed by two

peaks centred around 380, 362 and 400 8C (for the first maxima

peaks) and 476, 495 and 464 8C (for the second maxima peaks),

for (Ni15Pt0.1)BC, (Ni15Pt0.4)BC, and (Ni15Pt1)BC catalysts,

respectively. Furthermore, the Ni17Pt0 catalyst displays a

hydrogen consumption in the temperature range 260–375 8Cand a very large peak centered at about 425 8C, whereas the

other samples with 17% Ni show almost identical profiles with

a steady increase in the base line, followed by a large low-

temperature peak and a high-temperature peak. It must be

emphased that with a rise in Pt content till 0.4%, the first peaks

maxima where shifted toward lower temperatures, while the

second ones toward higher values in all catalysts whereas the

opposite effect is observed when the Pt amount was rised from

0.4% to 1%, these phenomena were more pronounced at higher

nickel loadings (Table 2).

In a previous paper [31], we have reported that, in the case of

Pt-free catalysts (NixPt0 samples), TPR profiles appear as

composite profiles of the corresponding monometallic ones,

Page 6: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Table 2

TPR peak maxima and optimum calcination temperatures for the (NixPty)BC catalysts

Catalyst Optimum calcination

temperature (8C)

TPR peak maxima and increase in base line

Increase in the base line (8C) First peak or shoulder (8C) Second peak or shoulder (8C)

Ni12Pt0 525 280–370 420 460

(Ni12Pt0.1)BC 500 277–371 400 471

(Ni12Pt0.4)BC 475 276–370 385 477

(Ni12Pt1)BC 465 278–370 405 465

Ni15Pt0 500 265–375 410 440 (shoulder)

(Ni15Pt0.1)BC 450 255–355 380 476

(Ni15Pt0.4)BC 440 25–335 362 495

(Ni15Pt1)BC 436 264–372 400 464

Ni17Pt0 450 260–375 425 (large peak) –

(Ni17Pt0.1)BC 360 167–247 321 549

(Ni17Pt0.4)BC 350 167–231 300 564

(Ni17Pt1)BC 300 239–354 396 526

Fig. 4. TPR profiles for (a) (Ni12Pty)BC, (b) (Ni15Pty)BC, and (c) (Ni17Pty)BC samples.

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111108

Page 7: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111 109

although somewhat modified by small shifts of the tungsten

species reduction peak toward lower temperatures and of the

nickel species reduction peak toward higher temperatures. We

have also reported that the increase in the base line was due to

the reduction of tungsten oxide species of various reducibilities,

while the low-temperature peak was ascribed to the reduction of

the nickel oxide and the high-temperature peak was assigned to

a further reduction of the tungsten oxide species. In the case of

Ni17Pt0, the large peak was the result of the superimposition of

two reduction peaks, the first of the NiO and the second of the

WOx species. Based on these observations and on the maxima

of TPR peak values given in Table 2, we can say that, in the case

of Pt-promoted catalysts ((NixPty)BC materials), increasing the

nickel loading induces a shift in the temperature corresponding

to the NiO peak toward lower values. As mentioned by Lucas

et al. [34], in the case of Pd–Pt beta agglomerated zeolite-based

catalysts, a lower reduction temperature could be associated to

the reduction of larger particles and/or particles that interact

weakly with the support, whereas a higher reduction

temperature could be associated to the reduction of nickel

that interacts strongly with the support. Larger nickel particles

were formed when the nickel loading increased. Dissociative

H2 chemisorption is easier on these large particles. Since H

atoms are better reducing agents than H2 molecules, the large

particles formed at high nickel content are reduced at lower

temperatures. The same feature was observed for platinum in

the concentration range 0.1–0.4%, the peaks representing the

reductions of nickel oxide shifted toward lower temperatures,

whereas those representing WOx species reduction shifted

toward higher values at higher platinum loading. However, it

should be emphasized that, in the case of (Ni17Pty)BC catalysts,

the introduction of Pt markedly decreased the temperature of

the nickel oxide reduction peak from 425 to 300 8C, which

suggests that the Pt effect is more pronounced at higher nickel

loadings. From the literature, it is well known that the presence

of nickel ameliorates the tungsten dispersion in solids [35],

besides and as earlier mentioned, the interaction between

platinum and tungsten is more pronounced when these species

are well dispersed, thus in the case of (Ni17Pty)BC catalysts,

where the tungsten is well dispersed, the strong interaction

between platinum and tungsten will weaken the nickel–

tungsten interaction and consequently the nickel reduction peak

will be shifted toward lower temperatures. In addition and as

above-mentioned, increasing Pt content from 0.4% to 1% result

in shifting the lower-temperature peaks (NiO peaks) toward

higher temperatures and the higher-temperature peaks (WOx

peaks) toward lower values. This effect may be interpreted by a

nickel–tungsten and platinum–tungsten interaction synergism;

at higher Pt concentrations Pt–WOx interaction is so weak that

Ni–WOx interaction effect will be more strong and more

noticeable, this latter is translated by increasing NiO peaks and

by lowering WOx species peaks

As mentioned by Wong et al. [36], Platinum reduction

usually occurs at a temperature between 100 and 200 8C, thus

we should mention here that no distinct peak due to platinum

reduction was observed in our platinum-loaded catalysts. The

same behavior was observed over Pt/ZrO2 calcined at 800 8C

[37]; it has been explained by considering that during

calcination, platinum oxide is decomposed to Pt0, detected

by XPS [37,38]. The phenomenon was previously found in

PtO2, which eliminates oxygen by heating above 350–450 8C[39]. On the other hand, in the case of Pt/SZ catalysts, Ebitani

et al. [40] and Shishido et al. [41] have demonstrated, using

XPS and EXAFS techniques, the presence of cationic Pt atoms

with poor or null properties for hydrogen chemisorption and

hydrocarbon hydrogenation–dehydrogenation [42], which

claim that Pt cannot be completely reduced to Pt0. Based on

these results, they proposed a layer of oxidized species on the

surface of a Pt0 core as a model for the Pt particle supported on

SZ. Besides, Contreras et al. [43] reported that the addition of

W6+ to 0.3% Pt/Al2O3 modifies both reducibility and

chemisorption capacity of Pt, presumably due to a strong

interaction with the oxide precursors, they stated that WOx

species produce changes in electronic and geometrical

properties of Pt particles in Pt/WO3–Al2O3 [44]. At their turn,

Yori et al. [45], mentioned that, in the case of Pt/WO3–ZrO2

catalysts, the presence of W affects Pt reducibility, the

phenomenon was explained by an WO3–Pt interaction.

Furthermore, Regabulto et al. [46] reported a decrease in CO

chemisorption upon introduction of W over Pt/SiO2. It was

argued that decoration of platinum crystallites by WOx moieties

was causing a partial physical blockage of the metallic sites.

Hoang-Van and Zegaoui [47] also found, using platinum on

WOx (for reduction temperatures in the 200–400 8C range) that

tungsten suboxides are formed at the Pt–WOx interface. These

suboxides can migrate onto platinum particles, preferentially

near the support surface. In addition, Larsen et al. [48]

considered the formation of mixed metal or metal oxide

clusters, to explain the low CO chemisorption in Pt/WOx–ZrO2

when platinum was added before calcination.

3.3. Effect of reduction temperature

The effect of the catalysts reduction temperature on their

activities was studied at 250 8C, at the optimum calcination

temperature and over a duration of 100 min. As shown in

Fig. 5a–c, regardless the Ni and Pt concentrations, the catalytic

activity increased with increasing the reduction temperature to

reach a maximum or a plateau and then decreased to remain

constant. Reduction at 600 8C yielded the lowest activity this is

believed to be due to the fact that higher reduction temperatures

induce a larger extent of undesirable tungsten (and support)

reduction [49]. Besides, increasing Pt content to 0.4% shifted

the optimum reduction temperature toward the lower values,

while the opposite trend was observed with a further rise in Pt

amount (Pt = 1%). These effects being more pronounced when

the nickel content was higher. In the case of Pt-free samples, it

was reported that the phenomenon was related to the presence

of reduced tungsten oxide species generated by hydrogen

reduction, these reduced species being more active than the

tungsten trioxide (WO3) [10,11,50,51] and to the tungsten–

nickel interaction, whereas in the case of Pt-promoted catalysts,

the phenomenon may be also interpreted by the presence of

WOx species which are more active than WO3 and by the

Page 8: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Fig. 5. Effect of reduction temperature on the performance of (a) (Ni12Pty)BC, (b) (Ni15Pty)BC, and (c) (Ni17Pty)BC catalysts. Reaction temperature = 250 8C,

TOS = 100 min.

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111110

synergism between Ni–WOx and Pt–WOx interaction. At lower

platinum loading the Pt–WOx interaction is very strong,

especially for solids with 17% Ni where the tungsten is well

dispersed, and it is well known that the Pt–WOx interaction

depends on the platinum and tungsten particle size, the smaller

the particle size the stronger the Pt–WOx intercation and

consequently the lower the Ni–WOx interaction. We have

previously reported that the latter effect (Ni–WOx interaction)

was translated by NiO reduction peak shift toward higher

temperatures, thus if this interaction is weakened, the NiO

reduction peak will shift toward lower temperatures and

consequently do the optimum reduction temperature. The same

result was reported by Larsen and Petkovic [49] with platinum

supported on tungstated zirconia catalysts and by Hua and

Sommer [52] with alumina-doped Pt/WOx/ZrO2 catalysts.

4. Conclusion

In the present study, we focused on the effect of the

pretreatment conditions on the catalytic performance of the Pt–

Ni–WOx/ASA catalysts prepared by a sol–gel method. The

collected data show that the catalytic activity of these solids is

strongly dependent on their calcination and reduction

temperatures. The most outstanding observations are as

follows:

� T

he catalysts activities depend strongly on their calcination

temperature; the incorporation of Pt shifts the optimum

calcination temperature to lower values.

� T

he density of acid sites is strongly influenced by the catalyst

calcination temperature, it increases to reach a maximum and

then decreases. Besides, the solids catalytic performance is

totally related to their acidities, the higher the sample acidity

the higher its activity.

� T

he reducibility of Pt–Ni–WOx/ASA (NixPty)BC samples do

not exhibit the PtO reduction peak, this behavior was ascribed

to the platinum–tungsten interaction, which is dependent on

the pretreatment conditions.

References

[1] M.H. Jordao, V. Simoes, D. Cardoso, Appl. Catal. 319 (2007) 1.

Page 9: Effect of pretreatment conditions on the catalytic performance of Ni–Pt–W supported on amorphous silica–alumina catalysts: Part 1. Catalysts prepared by a sol–gel method

Y. Rezgui, M. Guemini / Applied Catalysis A: General 335 (2008) 103–111 111

[2] S.D. Rossi, G. Ferraris, M. Valigi, D. Gazzoli, Appl. Catal. 231 (2002)

173.

[3] V.M. Benıtez, C.R. Vera, C.L. Pieck, F.G. Lacamoire, J.C. Yori, J.M. Grau,

J.M. Parera, Catal. Today 107–108 (2005) 651.

[4] S.R. Vaudagna, R.A. Comelli, N.S. Figoli, Catal. Lett. 47 (1997) 259.

[5] M.G. Falco, S.A. Canavese, R.A. Comelli, N.S. Figoli, Appl. Catal. 201

(2000) 37.

[6] A. Katrib, F. Hemming, P. Wehrer, L. Hilaire, G. Maire, J. Electron.

Spectrosc. Relat. Phenom. 76 (1995) 195.

[7] J.G. Santiesteban, D.C. Calabro, W.S. Borghard, C.D. Chang, J.C. Vartuli,

Y.P. Tsao, M.A. Natal-Sabtiago, R.D. Bastian, J. Catal. 183 (1999) 314.

[8] V. Logie, P. Wehrer, A. Katrib, G. Maire, J. Catal. 189 (2000) 438.

[9] A. Katrib, D. Mey, G. Maire, Catal. Today 65 (2001) 179.

[10] C. Bigey, L. Hilaire, G. Maire, J. Catal. 198 (2001) 208.

[11] C. Bigey, L. Hilaire, G. Maire, J. Catal. 184 (1999) 406.

[12] C. Bigey, G. Maire, J. Catal. 196 (2000) 224.

[13] S.V. Filimonova, A.V. Nosov, M. Scheithauer, H. Knozinger, J. Catal. 198

(2001) 89.

[14] C. Martin, G. Solana, P. Malet, V. Rives, Catal. Today 78 (2003) 365.

[15] K. van der Wiele, P.J. van der Berg, in: C.H. Bamford, C.F.H. Tipper

(Eds.), Comprehensive Chemical Kinetics: Complex Catalytic Processes,

Elsevier, Amsterdam, 1978.

[16] S. Maciej, K. Zbigniew, Chem. Eng. J. 90 (2002) 89.

[17] A. Scholz, B. Schnyder, A. Wokaun, J. Mol. Catal. 138 (1999) 249.

[18] C.U.I. Odebrand, A. Bahamonde, P. Avilar, J. Blanco, Appl. Catal. B 5

(1994) 117.

[19] P. Patrono, A. La Ginestra, G. Ramis, G. Busca, Appl. Catal. 107 (1994)

249.

[20] L.R. Pizzo, C.V. Caceres, M.N. Blanco, Catal. Lett. 33 (1995) 175.

[21] D. Kim, M. Ostromecki, I.E. Wachs, D. Kohler, J.G. Eckerdt, Catal. Lett.

33 (1995) 209.

[22] J. Leyrer, R. Margraf, E. Taglauer, H. Knozinger, Surf. Sci. 201 (1988)

603.

[23] J. Kijenski, A. Baiker, M. Glinski, P. Dollenmeier, A. Wokaun, J. Catal.

101 (1986) 1.

[24] M. Schraml-Marth, A. Wokaun, A. Baiker, J. Catal. 124 (1990) 86.

[25] U. Scharf, M. Schraml-Marth, A. Wokaun, A. Baiker, J. Chem. Soc.

Faraday Trans. 87 (1991) 3299.

[26] P. Iengo, M.D. Serio, V. Solinas, D. Gazzoli, G. Salvio, E. Santacesaria,

Appl. Catal. 170 (1998) 225.

[27] M. Campanati, G. Fornasari, A. Vaccari, Catal. Today 77 (2003) 299.

[28] M. Signoretto, M. Scarpa, F. Pinna, G. Strkul, P. Canton, A. Benedetti, J.

Non-Cryst. Solids 225 (1998) 178.

[29] F. Schmidt, Appl. Catal. 221 (2001) 15.

[30] M. Guemini, Y. Rezgui, A. Tighezza, A. Bouchemma, Can. J. Chem. Eng.

82 (1) (2004) 184.

[31] Y. Rezgui, M. Guemini, A. Tighezza, A. Bouchemma, Catal. Lett. 87 (1–

2) (2003) 11.

[32] Y. Rezgui, M. Guemini, Appl. Catal. 282 (2005) 45.

[33] M. Hino, K. Arata, Appl. Catal. 169 (1998) 151.

[34] A. Lucas, P. Sanchez, F. Dorado, M.J. Ramos, J.L. Valverde, Appl. Catal.

294 (2005) 215.

[35] D. Ouafi, F. Mauge, J.C. Lavalley, E. Payen, S. Kasztalan, M. Houari, J.

Grimblot, J.P. Bonnelle, Catal. Today 4 (1988) 23.

[36] S.T. Wong, T. Li, S. Cheng, J.-F. Lee, C.-Y. Mou, Appl. Catal. 296 (2005)

90.

[37] E. Ogata, Y. Kamiya, N. Otha, J. Catal. 29 (1973) 296.

[38] A. Dicko, X. Song, A. Adnot, A. Sayari, J. Catal. 150 (1994) 254.

[39] S.E. Livingstone, The Chemistry of Ru, Rh, Pd, Os, Ir and Pt, Pergamon

Text in Inorganic Chemistry, vol. 25, Pergamon Press, Oxford, 1973, p.

1334.

[40] K. Ebitani, H. Konno, T. Tanaka, H. Hattori, J. Catal. 143 (1993) 322.

[41] T. Shishido, T. Tanaka, H. Hattori, J. Catal. 172 (1997) 24.

[42] K. Ebitani, J. Tsuji, H. Hattori, H. Kita, J. Catal. 135 (1992) 609.

[43] J.L. Contreras, G. Del Toro, Y. Schifter, G.A. Fuentes, Stud. Surf. Sci.

Catal. 38 (1988) 51.

[44] J.L. Contreras, G.A. Fuentes, in: J.W. Hightower, W.N. Delgass, E. Iglesia,

A.T. Bell (Eds.), Proceedings of 11th International Congress on Catalysis,

Stud. Surf. Sci. Catal. 101 (1996) 1195.

[45] J.C. Yori, C.L. Pieck, J.M. Parera, Appl. Catal. 181 (1999) 5.

[46] J.R. Regabulto, T.H. Fleish, E.E. Wolf, J. Catal. 107 (1987) 114.

[47] C. Hoang-Van, O. Zegaoui, Appl. Catal. 164 (1997) 91.

[48] G. Larsen, E. Lotero, S. Raghavan, T.D. Parra, C.A. Querini, Appl. Catal.

139 (1996) 201.

[49] G. Larsen, L.M. Petkovic, Appl. Catal. 148 (1996) 155.

[50] A. Katrib, V. Logie, N. Saurel, P. Wehrer, L. Hilaire, G. Maire, Surf. Sci.

377–379 (1997) 754.

[51] A. Katrib, V. Logie, M. Peter, P. Wehrer, L. Hilaire, G. Maire, J. Chim.

Phys. 94 (1997) 1923.

[52] W. Hua, J. Sommer, Appl. Catal. 232 (2002) 129.