Influence of the nature of the support on the reducibility and catalytic properties of nickel: evidence for a new type of metal support interaction

Influence of the nature of the support on the reducibility and catalytic properties of nickel: evidence for a new type of metal support interaction

287 Applied Catalysis, 19 (1985) 287-300 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands INFLUENCE OF THE NATURE OF NICKE...

719KB Sizes 0 Downloads 10 Views

287

Applied Catalysis, 19 (1985) 287-300 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

INFLUENCE

OF THE NATURE

OF NICKEL:

EVIDENCE

P. TURLIER, Institut

ON THE REDUCIBILITY

FOR A NEW TYPE OF METAL SUPPORT

H. PRALIAUD,

de Recherches

1'Universite Cedex,

OF THE SUPPORT

Claude

P. MORAL,

G.A. MARTIN

sur la Catalyse,

Bernard,

Lyon

PROPERTIES

INTERACTION

and J.A.

Laboratoire

1, 2 avenue

AND CATALYTIC

DALMON

Propre

Albert

du CNRS, associe

Einstein,

69626

a

- Villeurbanne

France.

(Received

6 March5

1985, accepted

2 August

1985)

ABSTRACT Various nickel catalysts deposited on classical supports (SiO2, A1203, MgO) and on less conventional materials (Ti02, Th02, Ce02, ZrOp, Cr203) were prepared under conditions in which the strong metal support interaction of the Pt/Ti02type does not occur. Their catalytic properties for ethane hydrogenolysis and carbon monoxide hydrogenation and the reducibility of the nickel phase deduced Activity toward ethane hydrogenolysis from magnetic measurements were investigated. per unit area varies by two orders of magnitude when the nature of the support is changed. A correlation between activity and nickel reducibility is demonstrated and interpreted in terms of geometric effects of dilution due to the presence of unreduced surface-nickel species. These dilution effects are shown to play an important role in the case of the CO + H2 reaction. However, additional effects leading to changes in apparent activation energy (not yet well understood but possibly linked to the carbon monoxide coverage) result in a poorer correlation between activity, selectivity toward C2+ and nickel reducibility. These correlations lead to an interest in designing further experiments to study what can be termed the redox metal support interaction, which should not be confused with the strong metal support interaction.

INTRODUCTION The main role of supports form for high thermal of the supposedly catalyst.

is to carry the metal

stability.

inert carrier

exerts

the acidic

The possibility particularly

that the carrier

through

[I] and Solymosi

directly

its semiconducting

[Z]. More recently,

characterized

by a loss of chemisorption

activity

[5,6] probably into the metal

support

due to the migration phase.

areaffected

but, as stated

by Bond [7], "we are still the catalytic

properties

with

interaction, [3]. It is

compounds

the carrier

0 1985 Elsevier Science Publishers B.V.

from the properties

is acknowledged,

of our ability

the chemical

support". 0166-9834/85/$03.30

by Schwab

that the catalytic

in the infancy through

reactions.

phase,

support

of some reduced

by interaction

in the

[3,4] and a drop in catalytic

By now, the possibility

metal

and direct

the metallic

has been outlined

has been reported

capacity

of the whole

disclosed

in the reforming

type of metal

interaction,

of a supported

modify

influences

dispersed

that the nature

on the behaviour

properties,

another

strong metal

in a highly

however,

Pt/Si02-A1203,

support was shown to participate

the so-called

support

an influence

This is the case of the bifunctional

1950's where

particles

It has been recognized,

to control,

environment

of the

288 In this laboratory, interaction,

which probably

interaction toward

we have obtained

on various

supports nickel

according

to the nature

It seemed

the extension

paper, we report nature

to shed more

the metal

ascribed over

support

of activity

and selectivity

(Ni) is deposited

to the presence

the active

of

surface,

light on this type of support

study to a larger

number

of this study to another

the results

nickel

when

concentration

phase, was suspected

of the support

of supported

reaction

support

of the support.

the preliminary

to the metal

the strong metal

by changes

was tentatively

of variable

to us of interest

and to extend Furthermore,

patches

of a new type of metal

to do with itself

hydrogenation

[S]. This effect

unreduced

specific

has nothing

(SMSI) and which manifests

the carbon monoxide

evidence

of a more

systematic

in carbon monoxide

reaction,

to yield

on the reducibility

of supported

more

In this of the

and on the catalytic

hydrogenation

catalysts.

such as hydrogenolysis,

information.

study of the influence

of nickel,

effect

nickel

and ethane

behaviour

hydrogenolysis.

MATERIALS Unsupported first

Ni(OH),

[9] COnSiStS

removing

ammonia

precursors

in preparing

nickel

duction

for 2 h in a hydrogen

probably serious

prepared

at 300 K. This method

m*g'of

respectively.

were

a solution

after

decomposition

drawback

leads to a nickel

in a flow of helium

nickel

powders

of the glass apparatus.

for a model

catalyst

"unsupported"

and adding

136 m'/g)

(BET area, acetone

raising

flow of 4 l/h for 15 h, the BET surface

comparable

to the previous

designated

Ni- eSiO2 and unsupported nickel

used for Ni/SiO2, characteristics

Catalysts

hexammine Ni/Ti02,

by re-

(10.7 m2/g).

were obtained

Ni/A1203

This

of nickel

another

by processing nitrate

the temperature

hexammine

at 2 K/min

area of the nickel

The catalysts

is a

powder

thus obtained

in is

are

Ni, respectively.

according

are summarized

up to 0.2% silica,

and we have prepared

to the solution

at 690 K by linearly

one

containing

free from silica

a hydrogen

nitrate

of 12, 10, 7 and 4

attack

at room temperature

Supported

powder

The

and

for 1 h followed

yields

Ni(OH)*

of nickel

hexammine

from the ammonia

type of unsupported

[IO]. After reduction

to two techniques.

nitrate

gas flow of 4 l/h at 520, 570, 670 and 770 K,

This preparation

originating

according

of nickel

by adding

to a method

and Ni/MgO

in Table

the support

described

to a Solution

elsewhere

and already

[I']. Their morphological

1.

METHODS Reduction

was carried

out in a flow of hydrogen

those used for pure unsupported The degree the saturation according

of reduction magnetization

to a method

magnetization

catalysts

of the outgassed described,

reduced

under

conditions

similar

to

catalysts.

of the nickel

previously

of the sample

nickel

with

was obtained

samples consisting

the specific

by measurement

(10T6 Torr,

in comparing saturation

of

700 K for 2 h) the saturation

magnetization

of

289

TABLE 1 Morphological characteristics of supported precursors Support

Origin of the support

A1 203 Ti02 Si0 2

Degussa Oxid C Degussa P.25 Degussa A ZOO

CrZ03 Mg(OH)Z ZrO Z ThO Z Ce0 2

Degussa MgCl + NH z 40H ZrC1 + NH 4 40H Oxalate (NH 4)2 Ce(N03)6

BET aree /m 2 9 -1

Ni /wt%

Ref.

100 50 200

21 14 Z4

45 110 30 10 27

15 17.1 Z4 17.3 14.7

[11,12] [12,18] [12,13, 17] [19] [12] [14] [15] [16]

pure nickel [10]. Hydrogen chemisorption was used to evaluate the surface average diameter of nickel particles, DH" The diameter DH, which takes account of the accessible metal surface, is calculated on the following basis: surface stoichiometry at saturation, H/Ni S = 1; (100) and (111) planes on the surface; 8 = 0.65 H at 300 K and 10 Torr of hydrogen, as suggested by hydrogen isobars on Ni/Si0 [20]; 2 spherical shape of the nickel particle. Surface average diameters, OM' were also obtained from magnetic measurements in an electromagnet providing moderate fields (21 kOe) at 300 and 77 K, particularly for the case of superparamagnetic samples [10]. Some samples were also examined in a transmission electron microscope, which yields the surface average diameter DTEMwhen nickel particles are easily distinguished from the support (this is not the case for Ni/Th0 samples, for instance). 2 Catalyti c activiti es were measured at atmospheri c pressure usi ng a fi xed-bed reactor in dynamic and low conversion conditions. The effluents were analysed in a gas chromatograph equipped with flame-ionization and catharometric detections. For the case of carbon monoxide hydrogenation, activities and selectivities were measured under the following conditions: H2, 450 Torr; CO, 120 Torr; He, remainder; T, 525 K. The selectivity SC2+ is defined as 1-S Where S, is the selectivity 1 towards methane; the selectivity towards carbon dioxide is negligible. The activity towards ethane hydrogenolysis was measured under the following conditions: H 2, 136 Torr; C2H6, 8.1 Torr; He, remainder; total gas flow, 150 ml min- 1; T, 507 K. Intrinsic activities are expressed per unit area of metal accessible to hydrogen for both reactions.

290

TABLE 2 Diameters of nickel particles and apparent activation energies of ethane hydrogenolysis Samp1es

Reduction temperature /K

°H

OM

DTEM

Inm

Inm

Inm

673 773 873 973

4.1 4.5 4.7 5.3

5.4 6.7 7.2 8.1

4.7 6.2 7.2 8.1

Ni/A1 203

573 773 1073

2.4 5.6 8.5

4.1 6.0 8.1

Ni/Cr 203

723

Ni/MgO

12

Ea at 507 K IkJ mol-1 162 171 168 175 165 156 166

6

162 157

1051

20 11

Ni/Si0 2

920 1120 1190

6.5 15 18

6.5

162 175 175

Ni/Ti0 2

577 673 779 880

44.7 65 98 157

15 16.0

180 171

Ni/Th02

475 567 891

540 831

Ni/Zr0 2

558 585 646

187 690

Ni/Ce02

523 913

13.5 710

168 162 162 100-400 12 13.8

154 149

RESULTS Morphological properties Some characteristics of the supported catalysts are summarized in Table 2 (some of the results have been published elsewhere [11-19J). The results presented in this table deserve some comments. There is a good agreement between diameters deduced from hydrogen chemisorption, magnetism and electron microscopy, as far as those samples not in the SMSI state are concerned. This is the case for Ni/MgO and Ni/A1Z03, whatever their reduction temperature, and for Ni/Si02 and Ni/Ce02 reduced at moderate temperatures.

291

400

300 FIGURE

1

Reducibility

A3

- Ni/A1203,v4

@S

Ni/Zr02,+

500 of different

- Ni/Cr20s,a)

600 nickel

catalystsol-unsupported

5 - Ni/Si02,06

Ni,XZ

- Ni/Ti02,m7

Ni/MgD,

- Ni/Th02,

9 - Ni/CeO2.

As shown in Figure 1, the reducibility the nature

700

of the support.

of the nickel

Surprisingly,

some supports

phase depends

strongly

on

seem to play a promoting

role in the reducibility of the nickel phase. This is the case for Ce02 and Th02,

and, to a lesser extent, This interesting

effect

Size effect:

(i)

(at the beginning can be tentatively

the reducibility

of the reduction),

for TiO2 and Zr02.

attributed

can depend

to various causes: .2+ precursor an the size of the NI

particle. (ii)

Basicity

sidered

of the support:

as basic

the above

after Tanabe,

[21]. They can trap protons

four supports

and displace

can be con-

the redox equilibrium

to the right

Ni2+ + H

2

2 H+ + Ni"

=

(iii) Back spillover atomic

hydrogen,

Obviously

probably

which would

other experiments

Conversely, bility.

of hydrogen,

are needed

chemical

interaction

and nearly

with Si02 to give basic nickel

silicate

and Ni(OH)2

The reducibility are, respectively:

can react with

sequences

measured

role toward

between

nickel

reduci-

[ZZ], and is

the nickel compound 2+ and Mg '+ have the

into MgO (Ni

the same ionic radius).

interact

into

this question.

by Schuit and Van Reijen

in fact, NiO can be dissolved

of oxidation

on the support

agent of Ni2+.

to clarify

play an inhibiting

been outlined

due to the possible

and the support:

could be activated

be the reducing

MgO, Si02 and A1203

This has already

same degree

which

silicates

Ni(OH)2

or NiO can

(talc and antigorite)

A1203 to generate when the degrees

or a nickel

a basic aluminate. of reduction

are 25 and 75%

292

Ce02 > Th02 > Ti02 > ZrO2, unsupported

Ce02 > Th02 > unsupported

For higher

degrees

> ZrO2 > Ti02,

of reduction,

It is also interesting one given temperature,

> Cr203 > Al203

Ni/Th02

to notice

varies

Cr203 > A1203

becomes

This order

roughly

of the glue between sintering. responsible

samples.

FIGURE

Catalytic

2

for ethane

hydrogenolysis

- Ni/TiO2;a7

Ethane

than Ni/Ce02.

of the metallic

phase at

> Zr02, Th02

sequence:

and the support

on the support

which

reducible, would

at 507 K, versus - Ni/MgO;n3

- Ni/Th02;@8

reduction

- Ni/A1203;V4

- Ni/Zr02;+9

the nickel

to the Ni 2+ nature prevents

nickel

the glue concentration

be larger

(A) per unit area accessible

(3)

the larger

It can be attributed

that are not easily

adhesion

activity

unsupported Ni;$2

Ol06

particle

catalysts

for the metal

reducible

the reducibility

the dispersion.

the metal

On nickel

reducible

(2)

in the order:

parallels

the smaller

(1)

> SiO2 > MgO

more

that the dispersion

MN, Al *03, SiO2 > Cr203 > Ce02, TiO2 > unsupported

reducibility,

> MgO > SiO2

than on easily

to hydrogen

adsorption

temperature. - Ni/Cr203;

a

5 - Ni/SiO2;

- Ni/Ce02.

hydrogenolysis

Figure adsorption)

2 shows the activities toward

ethane

(per unit area

hydrogenolysis

of nickel

of the supported

accessible

to hydrogen

and unsupported

catalysts

as a function a sequence

of reduction

from Figure

Two major

factors

activity:

size-effect

temperature.

2, since the behaviour

can be invoked

smaller

size in the samples

to hydrogen

Ni/A1203

to occur

negligible

at a size

Ni/Cr203,

NilSiO2,

Ni/Ti02

and Ni/ZrfJ2

per unit area of nickel

temperature

increases

the highest activity of the normal

[14,17,18].

at the lowest

activity.

accessible Thus,

reduction

The case of Ni/ThD2,

where activity is almost unchanged with increasing reduction

and where

(Figure

are probably

is the lower limit of the average

expressed

the reduction

as being representative

and Ni/MgO,

temperature clear

when

to consider

to another.

of hydrogenolysis

(Table 2).

of the activity

adsorption

it seems reasonable temperature

value which

studied

from one sample

for the change The size-effects

of the SMSI for Ni/Ce02,

in a decrease

varies

it is very hard to deduce

they have been shown

In fact

than 5 nm [23], a critical

The occurrence results

to account

and SMSI occurrence.

over the size range considered.

particle

At first glance,

there

is no evidence

of the SMSI behaviour,

is relatively

2). The following sequence for ethane hydrogenolysis can then be

proposed: Cr203 > Ce02, unsupported This sequence of active energy,

> ThO2, Ti02,

can probably

sites rather

Zr02, A1203

be interpreted

than by electronic

of the number

since the apparent

165 + IO kJ/mol

seems

to be independent

equal

to that of unsupported

Activities

(4)

in terms of variation

effects,

(Table 2) and virtually

Carbon monoxide

> Si02 > MgO

of the nature nickel

activation

of the support

(171 kJ/mol).

hydrogenation

and selectivities

of the catalysts

are reported

in Table

3. The

activities, a, are expressed in carbon monoxide molecules converted into hydrocarbons per second and per cm2 of active surface

(as deduced

from Table

2, using

D,,). The activity

towards

methane

Data of Table 3 correspond a virtually

complete

low as possible conditions,

to a reduction

reduction

to avoid

SMSI effects (Ni/Ce02)

Ti02 > Zr02 > ThO2 > unsupported

The selectivity

(Ni/Cr203)

variations

towards

extent

to 0.1 (Ni/Si02).

of the solids

> 0.95);

on the catalytic

Cr203,

to the nature It follows

to is as

[8]. Under these

catalyst

is ca. 60. The sequence

> A1205,

leading

this temperature

activity

of the most active

C2+ hydrocarbons,

according

are a, and a2+, respectively.

temperature

(reduction

the ratio of the activity

that of the least active

important

and C2+ hydrocarbons

(Ni/TiO2)

for the activity

to is:

(5)

> MgO > Si02 > Ce02

S2+ = a2+/(a, of the support

the sequence:

+ a,,),

presents

and ranges

from 0.6

294

TABLE 3 Activities (a, expressed in 1012 mOlecules of carbon monoxide converted per second per cm Z of nickel), selectivities S and apparent activation energies E in the a C2+ CO + HZ reaction (500 K, HZ/CO = 3.8). Ea a 2+ Catalyst Red. temp. /K a1 a 1 + a2 \ 2+ a 1+a2+ /kJ mol- 1 Unsupported Ni/MgO Ni/A1 Z03 Ni/Cr Z03 Ni/SiOZ Ni/TiO Z Ni/ThO Z Ni/ZrO Z Ni/CeO Z CrZ03, ZrO Z' Ti02

675 995 920 773 9Z0 550 570 585 523

> A1 Z03 > ThO Z'

16 3.8 7.1 5 4 Z6 17 15 0.72 unsupported

25.4 5.1 13.4 12.5 4.4 57.8 28.8 34.9 1. 03

> MgO,

Ce0 2

0.37 0.26 0.47 0.60 0.09 0.55 0.41 0.57 0.30

> Si0 2

155 163 142 150 118 125 176 138 159

(6)

Apparent activation energies Ea were calculated at ca. 500 K and vary from 118 ± 9 (Ni/Si0 to 176 ± 9 (Ni/Th0 kJ mol- 1. Let us recall that SMSI effects 2) 2) do not affect parameters S2+ and Ea [8J. DISCUSSION fz~ hydrogenolysis As reported above, the invariance of Ea in ethane hydrogenolysis when changing the support suggests that the variations in catalytic activity can be correlated with a variation of the number of active sites rather than with a change of mechanism. The sequence of activity cannot be interpreted in terms of particle size (see sequence 2), of reducibility or basicity of the support [8J. In contrast, it can be observed that sequences (1) and (2), dealing with nickel reducibility, parallels rather well with sequence (4) obtained in hydrogenolysis activity if one excepts Ni/Cr This suggests that (a) the decrease in activity corresponding to sequence 203. (4) is connected to the decrease in reducibility of the nickel phase, and (b) that as proposed by us earlier [8J, the unreduced nickel species probably still present in the less reducible systems could behave as inactive areas which would dilute n the active surface. (It is difficult to imagine that the Ni + ions are alone on the metal surface: they are probably associated with anions such as silicates, aluminates, etc., originating from the support). This proposal agrees with the observed changes in activity without any change in Ea' We do not have a definite explanation of the particular behaviour of Ni/Cr203, which could be related to some specific properties of Cr203, and which deserves further experimentation.

295 CO + H2 reaction The above-mentioned nickel

species)

the problem

dilution

effects

have to be considered

here appears

vary considerably,

dilution

exist which

do change

two groups,

corresponding

Ea < 147 kJmol-‘,

intricate:

could

effects,

be related which

to Ea below and above

(2) and (3) on the other,

to the fact that apart

do not alter

Ea. One can then tentatively

the sequences

of unreduced

reaction, however, 2 apparent activation energies

(5) on the one hand and

This discrepancy

from the above-mentioned

with the presence

also for the CO + H

to be much more

and sequences

look very different.

(connected

Ea, some other effects

try to separate

the solids

147 kJmol-' (median value).

into

For

are:

Reducibility:

Ti02 > Zr02 > A1203

> SiO2

(2') reduction

25::

Zr02 > Ti02 > A1203 > SiO2

(3') reduction

75 z

CO + H2 activity:

Ti02 > ZrOZ > A12D3 > SiOp

For Ea > 147 kJ mol-':

Reducibility:

CeO2 > Th02 > unsupported

> Cr203 > Mgo (2", 3")

CO + H2 activity;

Th02 > unsupported

Except activity

> Cr203 > MgO > Ce02

for Ni/Ce02,

and reducibility

of [83, concerning on Ni/Si02, other

which

appears

the origin

Ni/Zr02

supports

are present

can withdraw

ment with an IR spectroscopic species

under the conditions

unperturbed

reducible

catalysts.

much

study:

study

Ni" atoms)

later,

the correlation

in both groups. of activity extended,

In this picture, in a geometric

thus leading

This hypothesis

each group,

solids,

to the

Nix+ species

Ni" atoms

(in agree-

the C-O bond of the

on these electron-deficient

result

Then, the proposal

from the surrounding

Cl33 and, therefore,

between

in the CO + H2 reaction

inside

on the less reducible

electrons

adsorbed

Ni" phase,

better

can be roughly

of the reaction.

(Nix' and neighbouring active

be discussed

of the variation

and Ni/TiO2

of the present

which

carbon monoxide

will

(5")

Ni" atoms

these

is not broken

inactive.patches

dilution

to a lower activity

, connected with a geometric

effect

of the

for the less dilution

effect,

296

a=a, +a2+ 50

FIGURE

3

Hydrogenation

to hydrogen

adsorption

- unsupported

01

of carbon

Ni/Si02;06

versus

Ni;

X

restricted

with

to solids

number

variations

of active

Concerning

where

atoms

than that of methane way

(the smaller

varies

with

4 - Ni/Cr203;

activity

a

5 -

reducibility

As a matter

of fact,

[23] and, therefore, merely

proposed

is

changes the

in terms of the

formation

will decrease

effect,

the catalytic

data obtained

Though

activity

and activity

on silica

should

of the various is rather

supported

implies

of C2+ more vary in a and acti-

samples

as a

poor, a general

This proposal

Ni-Cu alloys

x Cu content

cata-

up of nearly

the selectivity

can be observed.

a(x) of an alloy with

is made

This hypothesis

the production

the higher

the correlation

proposal

1231 that on NilSi

formation.

and that selectivity

with the above

that the activity

effects

the dilution

3 represents

based on activity shown

be interpreted

we have already

of their selectivity.

in keeping

cannot

dilution

rapidly

trend

v

-I- 9 - Ni/Ce02.

values.

in the mechanism

as that for methane

similar

Figure

- Ni/ZrO2;

site for C2+ hydrocarbon

active

surface

that the above-mentioned

vity).

(SC;) at 500 K, H2/CO Q 3.8. 3 - Ni/Al203;

that the correlation

changes

in activity

(a) per unit area accessible

sites.

lysts the surface

function

Activity

E, have comparable

with

the selectivity,

twice as many

n

- Ni/ThO2;@8

the observation

in Ea can be connected observed

selectivity

2 I Ni/MgO;

- Ni/Ti02;(a7

is consistent

monoxide.

is mainly

[23], which

(x = Cu/(Cu

have

+ Ni) at)

x:

CH4 formation

a,(x)

= a, (x=O)C1

-

xlN

(I)

297 with

N Q 12. Let us recall

nickel

atoms

forming

that in our hypothesis

the active

1231, N is the number

of adjacent

site.

C2-C4 formation

a(2_4)(x)

= a(*_4)

(C4+ products

(x=0) [l -

2N

xl

were observed

(1’)

in too small quantities

to extend

the relationship).

Then:

a2+

(x)

=K

a'

(x)

Relation line going

Cl - xl N = K'a, (x)

(1) should through

be observed

the origin

(2)

if only dilution

should

be obtained

effects

FIGURE 4 Hydrogenation of carbon monoxide. Ratio a2+/a, I" 0 Ni) at 525 K, H2/CO 2, 3.8. in lO'[ molecules/s/cm‘ 01:

Ni/Si02;a2,3,4:

atomic

ratios;a5:

In fact, supported

Figure

(the discrepancy

in activity

respectively

4 shows that relation where

versus

against

a,.

a, (a, activity

0.1, 0.2, 0.3 Cu/Ni

for C2+ formation

a silica-supported

[24] were

effects

is probably

is a little K-promoted

shown to originate

from Ni-Cu solids

(2) is obeyed

pure dilution

from a linear correlation

ensemble

On the contrary,

with

i.e. a straight

+

cu

Ni-K/Si02.

Ni-Cu catalysts

calculated

different

Ni-Cu/Si02

occur,

for plots of a2+/a,

(Figure 4).

very well for sil ica govern the changes

due to the fact that the

less than twice

nickel

in activity

catalyst,

from an electronic

that for methane.

where

effect,

the changes

is clearly

V

FIGURE

5

in lOI

a1

I

I

10

20

Hydrogenation

of carbon

molecules/s/cm2

5a) A

: Ni/A1203; @

5b)O

: unsupported

work.

is roughly

Ni;

x

the most active is, however, if Ni/ThOp, Ni/Cr203

solids

with

(Ni/Ti02,

worse

unsupported

Ni/TiO2;

behaviour.

could

in activity

monoxide

pressure)

selectivity

species)

and selectivity. probably

depends

are then affected,

With Ni/Cr203, reducibility

for the variations

with

These

(2)

T_:' correlation

Ea > 147 kJmo1

(Figure

(2), Ni/Ce02

Ni/CeO2

and Ni/Cr203,

discrepancies

owing

species which

to the support.

to the kinetics

of the reaction:

"CO (coverage

role and governs

of the support,

to some scattering

is obtained

selectivity

in unthe

and carbon

activity

(Figure

when paralleling

Z", 3" and 5"), however,

some

could originate

diluting

law where

5b):

and particularly

As oCO (for a given temperature on the nature

in the

and

plays an important

thus leading

a good correlation

(see sequences

observed

5a), relation

on the surface

a kinetic

: Ni/Ce02

is the least selective

relation

be connected

we have proposed

monoxide

changes

: Ni/Cr203; +

such as the nature of the unreduced distributed

catalyst

v

Even excluding

for both series.

of disagreement

carbon

a, (a, activity

: Ni/Zr02

(Ni/Si02)

follow

Another

dissociated

@

ZrO2) are the most selective.

could be more or less randomly

on a Ni/Si02

versus

Ea < 147 kJmor'(Figure

for the catalysts

are observed

source

a2+/a,

Ni/ThO2;

solid

Ni and Ni/MgO

have a different

sources,

Ratios

(2) can account

the least active

clearly

discrepancies in different

q:

: Ni,JMgO:a:

For the catalysts

obeyed;

monoxide.

a1

I

20

Ni) at 500 K, Hz/CO s 4.

: Ni/Si02;

Now, let US see if relation present

1

10

and

5).

activity towards

and

C2+ is

299

unambiguously lytically

high when compared

active

functional-type towards

Cr203 where

The case of Ni/Ce02

however,

Ni/Ce02

This peculiar

size effects difficult

to put forward

not present

but needs

of Ni/Ce02

since other

a bi-

from nickel

could occur.

has the lowest reduction

the most active of the series

trapping

in ethane

particle

species

are

with these same properties

(and decreasing studies

hydro-

in the CO + H2

carbon monoxide

interaction

Further

support.

could migrate

cata-

has not yet been explained:

catalysts

Electronic

of the C-O bond

experimental

and one can imagine

this catalyst

and is accordingly

of the support

is known as being

via alkylation

is the least active

this behaviour.

to a reinforcement

chain growth

behaviour

and basicity

reactions

some CHx adspecies

is different,

of all the solids

genolysis, reaction.

in which

Cr203

solids.

in hydrocarbon-reforming mechanism

the support

temperature

to other

do

of nickel with

Ce02,

leading

the activity),

could

be proposed

are planned

to try to solve

this

point.

CONCLUSION the variations

In C2H6 hydrogenolysis, for nickel depends

can be correlated

strongly

higher the activity according

linked

to variations

reduced

dilution

reaction. energy

per metallic

correlating

surface suggests

in the number species

effects

However,

(probably

of the support.

to the carrier

surface-nickel

These

with the reducibility

on the nature

energy

of the support

the field of catalysis:

the joint

various

in redox systems this joint

reactions

methanol

resulting

effects)

of the apparent

changes

in activity

due to the presence

are

of un-

role in the CO + H2

in changes

in apparent

activation

lead to some discrepancies

in the reducibility

which

presence

species

probably

when

of the supported

play an important

of reduced-unreduced

shown

is proposed

(enantioselective

in further

could

which were

presence

[28], oxygenated

an interest

the

and reducibility.

effect

Moreover,

phase, which

the reduction,

unit. The invariance

sites,

the support

act as a diluent.

effects

a kind of metal-support

can result

changing

of the metallic

The easier

also seem to play an important

additional

The interference

when

that the observed

of active

which

due to electronic

activity

of activity

capable

hydrogenation

[27],

can be dubbed

hydrogen

[26].

in the mechanism

selective

[29] from the CO + H2 synthesis)

study of what

is

metal species

of activating

as being necessary

phase

role in

formation

of

of

and leads to

the redox metal-support

inter-

action.

REFERENCES 1 2 3 4 5

G.M. Schwab, Adv. Catalysis, 27 (1978) 1. F. Solymosi, Cat. Rev.,1 (1967) 233. S.J., Tauster, S.C. Fung and R.L. Garten, J. Am. Chem. Sot., 100 (1978) P.G. Menon and G.F. Froment, J. Catalysis, 59 (1979) 138. G. Den Otter and F. Dautzenberg, J. Catalysis, 53 (1979) 116.

170.

300

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

24 25 26

27 28 29

a.H.

Ellestad and C. Naccache, Proc. 12th Swed. Symp. Catalysis, Lund, Oct. 11, 1979, CWK Gleerup, Lund, 1981, p.95. G.C. Bond, Studies in Surface Science and Catalysis, Vol. 11 "Metal-support and Metal-Additive effects in Catalysis", Elsevier (1982) 1. P. Turlier, J.A. Dalmon and G.A. Martin, Studies in Surface Science and Catalysis, Vol.ll "Metal-support and Metal additive effects in Catalysis", Elsevier, (1982) 203. A. Merlin and S.J. Teichner, Comptes-Rendus Acad. Sci., Paris, 236 (1953) 1892. M. Derrough, R. Bardet, P. Turlier and Y. Trambauze, "Les Solides finement divises ", Ed. Direction de la Documentation Francaise, Paris (1969) 47. P. Turlier and G.A. Martin, React. Kinet. Catal. Lett., 19 (1982) 27. G.A. Martin, N. Ceaphalan, P. de Montgolfier and B. Imelik, J. Chim. Phys . , 10 (1973) 1422. M. Primet, J.A. Dalmon and G.A. Martin, J. Catalysis, 46 (1977) 25. P. Turlier and G.A. Martin, React. Kinet. Catal. Lett., 21 (1982) 387. P. Turlier and G.A. Martin, React. Kinet. Catal. Lett., 25 (1984) 1. A. Maubert, G.A. Martin, H. Praliaud and P. Turlier, React. Kinet. Catal. Lett., 24 (1984) 183. G.A. Martin and J.A. Dalmon, React. Kinet. Catal. Lett., 16 (1981) 325. A. Choplin, J.A. Dalman and G.A. Martin, Comptes-Rendus Acad. Sci. Paris, 293 (1981) 137. H. Praliaud and G.A. Martin, Studies in Surface Science and Catalysis, Vol.17, "Spillover of Adsorbed Species", Elsevier (1983), 191. G.A. Martin, J. Catalysis, 60 (1979) 345, 452. K. Tanabe, Studies in Surface Science and Catalysis, Vo1.20, "Catalysis by acids and bases", B. Imelik et al., Eds., Elsevier (1985) 1. G.C.A. Schuit and L.L. Van Reijen, Adv. in Catalysis, 10 (1958) 242. G.A. Martin, J.A. Dalmon and C. Mirodatos, Proc. 8th Int. Congress Catalysis, Dechema 1984, IV, 371; J.A. Dalmon and C. Mirodatas, J. of Malec. Catalysis, 25 (1984) 161; J.A. Dalmon and G.A. Martin, Proc. 7th Int. Congress Catalysis, Kodanska (1981),402; J.A. Dalmon and G.A. Martin, Proc. 8th Int. Congress Ca ta 1ys is, 1984, 185. H. Praliaud, J.A. Dalman, C. Mirodatos and G.A. Martin, submitted ta J. Catal., J.A.Dalmon and G.A. Martin, J. Catalysis, 84 (1983) 45. R. Dutartre, P. Bussiere, J.A. Dalmon and G.A. Martin, J. Catalysis, 59 (1979) 382; J.A. Dalmon, C. Mirodatos, P. Turlier and G.A. Martin, Studies in Surface Science and Catalysis, Vol.17, "Spillover of Adsorbed Species", Elsevier (1983) 169. A. Hoek and W.M.H. Sachtler, J. Catalysis, 58 (1979) 276. E.K. Poels, R. Koolstra, J.W. Geus and V. Ponec, Studies in Surface Science and Catalysis, Vol.11, "Metal-support and Metal-additive effects in Catalysis", Elsevier, (1982) 233. P.R. Watson and G.A.Somorjai, J. Catalysis, 74 (1982) 282.