Report on the catalytic activity of 6-heteropolymolybdates as potential Fischer-Tropsch synthesis catalysts

Applied Catalysis, 21 (1986) 61-72 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

REPORT

ON THE CATALYTIC

TROPSCH

SYNTHESIS

ACTIVITY

61

OF 6-HETEROPOLYMOLYBDATES

AS POTENTIAL

FISCHER-

Ann Arbor,

MI 48109,

CATALYSTS

Jo-Wei DUN, Erdogan GULARI* and Barry STREUSAND' Department

of Chemical

Engineering,

University

of Michigan,

U.S.A. *to whom correspondence

should

be addressed

'Climax Molybdenum

Company,

(Received

1985, accepted

27 March

Ann Arbor,

MI 48106,

t October

U.S.A.

1985)

ABSTRACT Unsupported and supported decomposition products of 6-heteropolymolybdates were tested as possible Fischer-Tropsch synthesis catalysts. Depending on the heteroatom, the activity changes significantly and the best supported catalysts have higher activity (based on total metal loading) than previously investigated Moo2 and Mo2C catalysts.

INTRODUCTION Molybdenum denitration

catalysts

are extensively

and selective

oxidation

charcoal

supported

prepared

[3-51. The advantages

resistance

to sulfur

the ability with

activity

expense

yields

catalysts)

[3-51.

of a moderate

We have recently activities

heteropoly

most

commonly

Paper we report prepared

activity.

FT catalysts

distribution without

research

supported

are their

(C, to C,) and

coking.

These are balanced

products)

and (for most

has shown that both the

molybdenum

in C 2+ selectivity

the possibility

of molybdenum

of molybdenum

our results

Recent

were successfully

compounds

comes

can be

about at the

in activity.

investigating

used starting

formula

compound

catalysts

instead

of ammonium

in preparing

for both charcoal

of increasing

simultaneously

supported

both the

by starting

heptamolybdate,

supported

of the 6-heteropolymolybdates

X is the heteroatom

One example octahedron

based

product

hydrohigh activity

catalysts.

and unsupported

the

In this catalysts

from 6-heteropolymolybdates.

The general where

compounds

the narrow

of charcoal

and selectivities

with

of using molybdenum

the increase

decrease

begun

(FT) catalysts

(30 to 50 wt% of the hydrocarbon

moderate

However,

[1,2]. More recently

low H2 to CO feed ratios

and C2+ selectivity

increased

Fischer-Tropsch

poisoning,

to work with

high methane

molybdenum

molybdenum

used in hydrodesulfurization,

reactions

0166-9834/86/$03.50

(Co, Al, Cr, Fe, Rh, Ga or Ni).

of the structure

is surrounded

is: (NH4),_,[XMo60,,H6].jH20

of the anion

by six Moo6 octahedra

is shown

in Figure

in a planar

0 1986 Elsevier Science Publishers B.V.

1. A central

configuration.

The

X06

62

FIGURE

1

Diagram of the structure

of the 6-molybdo-chromate

with atoms drawn to the scale of their is chromium Open

ion, and the six solid spheres

spheres

ammonium

represent

oxygen

water of hydration

composition

begins at around

selectivity

of molybdenum

pretreatment

conditions

surrounding

of water

were reviewed is lost first.

Our goal in this research

(III) anion

[6]. The central

represented

solid sphere

it are molybdenum

ions.

ions.

salt also has seven molecules

all the 6-heteropolymolybdates heating,

ionic radii

of hydration.

in detail

The properties

by Tsigdinos

Constituent

water

of

[61. Upon

loss leading

to de-

473 K. was to find out if we could modify

catalyst

in FT synthesis

by varying

the activity

and

the heteroatom

and

for these 6-heteropolymolybdates.

EXPERIMENTAL Materials Ammonium

salts of 6-molybdo-nickelate,

were obtained further

from Climax

purification.

manufacturer,

this charcoal

pore size distribution 20 i in diameter. contains

with

low amounts

cobaltate

Labs and were

and aluminate

used without

was used as a support. According to the 2 -1 and a bimodal has a BET surface area of 1100 m g 50% of the pore volume

(
ferrate,

CPG charcoal

From the chemical

was used as a nonporous

Co., Ann Arbor

Molybdenum

Calgon

chromate,

analysis

being due to pores

supplied

of Fe, K, Ca and Na impurities.

high exterior

surface

larger

by the manufacturer

area support

Degussa

than

it

P-25 Titania

(BET surface

area of

50 m2 g-l) for one of our catalysts.

Catalyst

preparation

Details testing X-ray

and testing

of the standard

of their performance spectra

taken

procedure

for preparing

the supported

has been given elsewhere

initially

verified

catalysts

and

[5].

that all the catalysts

had the previously

63 reported

structures

[6]. In order

catalysts

under the reaction

catalysts

were also taken.

these samples

were allowed

and then exposed the unsupported

catalysts.

but quite

of strains. oxidation

by X-ray

oxidized

should

to decide

catalysts,

was converted

crystallite

line broadening

of cata-

only Moo2 was found to indicating

the presence

that some lower

For the charcoal

supported

size was found to be 200 + 40 i. For the

sizes were

(i.e. larger

too large to be reliably

measured

than 1000 i).

4

0

hydrogen

oxidation

the supported

in air we suspect

to Mo02.

and used

in flowing

whether

asymmetric,

in the

X-ray diffractometer,

lead to only surface

were taken

Mo02 crystallite

present

of the reduced

to a Philips

or not). To our surprise

Since the X-ray spectra

unsupported

phases

to cool down to room temperature

a few of the peaks were

the average

spectra

taken

(It is difficult

state MO compound

catalysts,

the actual

X-ray

conditions,

Before being

to air. This process

lysts were also surface be present

to identify

8

12

16

Time (Hr.) FIGURE

2

Self-reduction

Reaction

conditions:

of the moles

of CO reacted

Mo,Ni;

n

RESULTS

AND DISCUSSION

, MO-CO;

In our first loaded

lytic activity

heating

0,

attempt,

into the reactor

only two hours.

detectable

and activation

were

catalysts.

rates are in the units

of moly atom-min.

Symbols:

0,

Mo-Cr.

0.15 g of unsupported and tested

after

drying

conditions

was too low to produce

products

6-heteropolymolybdate

= 3. Specific

to form hydrocarbon/mole

At the reaction

the reactor

of several

573 K, 0.12 #Pa, H2/C0

CO2 and water,

6-molybdo-nickelate

in flowing

helium

amounts

= 3 the cata-

of hydrocarbons.

at a low conversion that the hydrocarbon

was

at 423 K for

of 473 K and a feed H2/C0

detectable

to 573 K we noticed

ammonium

of only

The only

0.2%. Upon

production

increased

c2+ selectivity

( wt%)

74 *

50

0

200

150

loo Time (Mitt)

FIGURE

3

Specific

rates of unsupported

rate units and the reaction All catalysts 0

, Mo-Cr;

with

are reduced 0,

time as shown

conditions

6-heteropolymolybdate

at 773 K for four hours.

MO-AI;

catalysts.

are the same as that given Symbols:

Specific

in Figure

n,

Mo-Ni;

2.

m,

MO-CO;

MO-Fe.

+,

in Figure

is due to the nickel

2. This, we believe,

catalyzed

self-reduction

of the catalyst. By activating the hydrogen, nickel catalyzes the +4 or possibly even lower oxidation states which may in turn of MO +6 to MO

reduction catalyze

hydrogenation

for several activation in Figure

peaking.

was also observed

with

On the other

at this temperature

The shape of the deactivation The initial

for aluminate nickelate

also

and ferrate.

and chromate

which

indicates

Table reduction

the cobalt

in activity

autocatalyzed

and chromium

deactivation

According

would

not shown

rates

heteropolies,

is highest

have approximately would

catalyst

steady

have roughly

treatments.

is not active

reaction

We see that after

have significant

activity.

reduction

catalysts.

of the particular

at these

rates

and-lowest indicated,

state activity.

the same steady and chromate.

On state

Figure

conditions.

rate for the bulk catalysts

After

no detectable

for the nickelate

deactivation

equal

as shown

in the figure.

is about one third of that of the nickelate that ferrate

showed

and

is in the

for the unsupported

to the initial

and aluminate

continued

reduction

for these heteropolies

curve seems to be a characteristic

1 shows the specific

heteropolies

increase

hand, Al and Fe heteropolies

rate of deactivation

the other hand, cobaltate activity

This

but slower

and are therefore

Figure 3 shows the initial

catalyst.

Similar

2. We see that the ease of reduction

order of Ni>Co>Cr. activity

of carbon monoxide.

hours before

after various

at 573 K only Ni, Co and Cr

four hours of reduction

at 673 K

3

65

TABLE

1

Specific

reaction

ratesa

of unsupported

6-heteropolymolybdates

in Fischer-Tropsch

synthesis

Specific

rateb

Reduction

Hetero

conditions

atom 573 K, 4 h

673 K, 4 h

773 K, 4 h

773 K, 12 h

Ni

0.0092

0.0089

0.0118

0.0288

co

0.0037

0.0232

0.0049

0.0087

Cr

0.0024

0.0168

0.0119

0.0055

0.0069

0.0031

0.0041

3.7 x 1o-4

9.3 x 1o-4

Al 8.7 x 1O-5 Fe

0.004.4 5.3 x lo+

aReaction

conditions

Rate values b

Specific

mole

573 K, 0.12 MPa, H2/C0

are taken at time equal

reaction

of moly

the activity

rate is defined

of the Ni-containing more

active.

and chromium

catalysts

supported

reported

for the bulk molybdenum

values

rates are quite

aluminate

high.

changes.

than twofold.

active.

These

tested

The Sncreases

carbide

[7-91 and 0.0040

respectively

qnd 0.0044

[IO]. Considering catalysts

reduction

these

to 773 K brings active,

respectively.

the chromate

the increased results

which

observed

with

processes reduction

in higher

specific about

chromate

The most dramatic

a 370% drop

to

for charcoal

is 30% more

except

reactiop

the fact

temperature

and decreases

both

0.0042

the reduction

based catalyst:

is increased

Increased

oxide,

based catalyst

and time are due to two opposing

as the temperature

treatment

Our specific

rates may be compared

for the bulk molybdenum

(5-10 m2 g-') unsupported

Increasing

catalysts

and has the units of: moles of CO

in activity.

time at 773 K from four hours to 12 hours results

for all the catalysts

sintering.

area

Nickelate

this reduction

are quite

of moly atom-min. of 0.0014

and carbide

by the cobaltate

the reduction

temperature

amount

show 30% and 55% drop in activity

is exhibited

to form hydrocarbon/

stays the same but all other

catalysts

molybdenum

oxide

that these are low surface

more drastic

of CO reacted

We note that after

containing

to form hydrocarbon/mole

the previously and 0.0029

to 2 h.

as the moles

rate is based on the bulk molybdenum reacted

= 3.

atom-min.

are significantly the cobalt

used:

decrease

Lengthening

in increased

shows a decrease the increased

and

rates

of more

reduction

going on simultaneously: is also accompanied

activity

while

sintering

by results

66 TABLE

2

Specific

reaction

ratesa

of 6% supported

6-heteropolymolybdates

in Fischer-Tropsch

synthesis

Specific

ratesb

Hetero Reduction

atom

conditions

573 K, 4 h

673 K, 4 h

Ni

0.0141

0.0162

0.0137

0.0123

co

0.0112

0.0238

0.0392

0.0142

0.0069

0.0137

0.0148

Cr

773 K, 4 h

773 K, 12 h

6.7 x 10 -4 Al

0.0011

Fe

0.0042

0.0105

0.0049

0.0107

0.0121

0.0168

0.0101

0.0231

0.0238

0.0095

0.0061

0.0118

2.9 x 1fY4 AHM 9.1 x 1o-4 cot 9.0 x 10-4

aReaction b Specific cMo-Co

conditions reaction

supported

should

used,

therefore

one to reduce.

trend of increased

activity

hand, the other

reduction

temperature

In order active

The nickelate

catalysts

area we prepared

increase

activity

also shows

temperature

problem

and to increase

We

is the

similar

and time.

activity

in

at all

in activity.

of all and clearly

based catalyst

reduction

the decrease

of reduction

increase

the maximal

6 wt%

(by molybdenum) reduction

can be seen by comparing

based catalystsupportingon after

reduction

decrease

in activity

with

activity

being

55% of the unsupported

about

degree

exhibit

rates at different

of the support

activity

based catalyst

in a continuous

has the lowest

the sintering

see that for the nickelate twofold

1.

On the

at medium

and time.

of reaction

2. The effect

in Table

up by the increased

at increasing

three

to decrease

surface

The results

1.

For the ferrate

resulting

note that this catalyst

most difficult

other

area catalyst.

area is more than made

temperatures

in Table

as that given

on titania.

in a lower surface the surface

used as that given rate is defined

increased

charcoal conditions Table

supported

results

in Table 1. We

in a

is a continuous

time and temperature,

catalysts.

the

catalysts.

are shown

2 with Table

charcoal

at 673 K but there

reduction

(if possible)

the maximum

This indicates

that sintering

67

Specitic Reaction Rate

150

loo

50

0

(*lCW

MO

Time (Min.)

FIGURE

C2 + selectivity

4

vity is defined

as nCn/nCn.

Figure

2. Symbols:

occurs

more

its maximum

maximum

is more

activity

supported

lyst prepared supported

is slightly

readily

reduced.

catalyst.

For comparison,

based catalyst.

catalyst

the importance

Figure

maximum

the higher

while

is

activity

ferrate

increases

based catalyst

one, with the

that the supported

based catalyst

ferrate

exhibits

the

we have also listed (AHM) solution

a supported

to E-heteropolies,

we find

after the 12 h reduction

to the charcoal

cata-

and a titania

supported

at 773 K.

catalyst,

of the support.

C2+ selectivity. nickelate

hydrocarbons

value

catalysts.

catalyst's

of all the unsupported

at 773 K. It is found that this reduction

in desired

the highest

and reaches

of 773 K. It is also the lowest

activity

is inferior

4 shows the C2+ selectivities

4 h of reduction

indicates

As a contrast

has its highest

supported

unsupported

to the unsupported

heptamolybdate

The titania

The cobaltate

temperature

The supported

The aluminate

after the 12 h reduction

that the AHM catalyst

showing

lower.

in

#:, MO-Fe.

at 773 K. This maximum

as compared

This clearly

MO-Al;

catalyst.

chromate

Selecti-

as that given

0,

reduction

of the corresponding

from a pure ammonium

cobaltate

at higher

catalysts.

conditions

Mo-Cr;

for this supported

in activity

being four times higher.

of sintering

0,

the supported

catalyst,

increase

and reduction

four h of reduction

but the maximum

based catalyst effect

after

6-heteropolymolybdate

MO-CO;

its activity

than the maximum

the unsupported

shows the most

m,

than reduction

activity

continuously

Reaction

Mo-Ni;

increases

70% higher

Unlike

0,

readily

based catalyst

about

of unsupported

We see that ferrate

based catalyst

about

with

after

gives the

based catalyst

gives the lowest

half is ethane,

catalysts

condition

exhibits

C2+ selectivity.

the balance

being propane

Of and

68

I

481

I

I

0

50

loo

150

200

Time (Min.)

FIGURE

5

C2+ selectivity

Selectivity in Figure

is defined 2. Symbols:

small amounts that, with

of 6% charcoal

as nCn/nCn. 0,

Mo-Ni;

of ethylene,

the exception

butane

supported

Reaction

n,

MO-CO;

0,

and pentane.

of the nickelate,

6-heteropolymolybdate

and reduction

conditions

Mo-Cr;

However,

0,

catalysts.

as that given

MO-Al;

*,

it is interesting

the variation

in selectivity

MO-Fe.

to note is only

f 10% among all catalysts. Figure in steady

5 shows the C2+ selectivity state C2+ selectivity

unsupported being

catalysts,

largely

for the more Reactivity

two thirds

propane. active

with

of the supported

of the C2 + fraction

Small amounts

of butane

in Table

3 and Table 4 for the activity nickelate

case the activities

obtained

the

were observed

only

using CO2 feed. The results

and C2+ selectivity based catalyst

from using CO2 feed of many methanation

(for example,

and selectivity

values

et al. [IS] are also

ref. [II-261).

for bulk molybdenum

included

of 673 K and four h. In

since previous

such as Ni, Fe, Ru and Co based catalysts

that of CO feed

respectively.

is'significantly

are far less than their corresponding

with CO feed. This is not surprising

the activity catalysts,

our catalysts

than that of using CO feed at the reduction

every other activities

and ethylene

Unlike

with the balance

CO2 feed

We note here that only the unsupported more active

is ethane

The variation

catalysts.

Instead of using CO, we also tested are shown

catalysts.

is only f 5% among all catalysts.

have shown

and Fischer-Tropsch are significantly

For comparison,

based catalysts

in these tables.

studies

Since these

literature reported

synthesis lower than activity

by Anderson

literature

values

are

69 TABLE

3

Specific

reaction

previously

ratesa

reported

of our t-heteropolymolybdates

bulk molybdenum Specific

catalysts reaction

using

ratesb

Reduction

Hetero-atom

catalysts

in comparison

with

CO2 feed *IO

+4

conditions 773 K, 12 h

673 K, 4 h

773 K, 4 h

Ni

195.0

26.6

25.2

co

24.5

11.1

12.1

Unsupported

Cr

2.7

6.8

9.3

Al

1.6

1.7

2.4

Fe

< 1.0

< 1.0

< 1.0

Charcoal Supported Ni

18.7

3.0

3.9

co

27.7

10.0

8.6

6.1

2.9

2.7

AHM

Reaction

at

Bulk catalyst 573 Kd

623 KC

from ref. [II] Moo3

0.0

0.0

Moo2

10.8

1.9

MoS2

1.8

0.3

MO

3.5

0.6

Mo2C

44.3

7.6

Mo2N

1.2

0.2

aReaction

conditions

bSpecific

reaction

'These d

are initial

used as that given rate is defined rate at reaction

in Table

as that given conditions

1. in Table

1.

of: 623 K, 1 ATM, H2/C0

= 3.7, from

ref. [Ill. These

rates are converted

activation

energy

from rates at 623 K to 573 K, assuming

an apparent

of 25 kcal mol -' for all catalysts.

taken at 623 K and not at 573 K, the temperature comparable

to ours.

For simplicity

reasonable

apparent

activation

in converting

energy

we used, they are not directly

their values we have assumed a of 25 kcal mol -1 for all catalysts they

70 TABLE 4 C2+ selectivitya ly reported

of our t-heteropolymolybdates

bulk molybdenum

catalystsb

catalysts

in comparison

with previous-

using CO2 feed.

C2+ selectivity Hetero-atom Reduction 673 K, 4 h

conditions

773 K, 4 h

773 K, 12 h

Unsupported Ni

3.0

32.6

co

3.5

34.3

33.0

Cr

7.7

27.8

31.8

Al

0.0

2.0

11.9

Fe

0.0

1.8

10.8

Ni

11.7

1.6

1.4

co

0.1

0.1

0.1

AHM

0.1

0.2

0.1

26.6

Charcoal Supported

Reaction

at

Bulk catalyst from ref. [11]

623 K

Moo3

0.0

Moo2

2.0

MoS2

0.0

MO

2.0

Mo2C

24.3

Mo2N

0.0

aC2+ selectivity as that given bLiterature

tested.

in Table

values.

(nickelate, magnitude

values

activity

are based on the initial

values

are listed

in Table

We see that, with the exception

cobaltate

more active

same activity

at time equal to 2 h.

Reaction

conditions

1.

C2+ selectivity

The converted

converted

values

are based on the values

and chromate

based catalyst)

than their bulk catalysts.

as our nickelate

show some interesting

of Mo2C,

and cobaltate

points.

values.

3 together

are in general

catalysts.

catalyst

an order of

Their Mo2C catalyst

based

with the un-

our more active

is about

the

The C2+ selectivity

We see that our more active

catalysts

exhibited

71 about

30% C2+ selectivity

lysts except favors

being reduced

Mo2C show a negligible

the formation

their catalysts considering

after

of methane

would

the negligible

would

have a comparable

lyst has the advantage

by Anderson dispersed

The role of the heterometal

were a separate

FTS activities

the overall

significantly

different.

for catalysts

with

to our catalysts.

lattice

they could there

having

activity

is one to six. Thus if the comparable

of our bulk catalysts

activities

atom

in the Moo2 lattice.

catalysts

and would

that the heterometal are activated

prepared

using pure AHM. On the opposite

catalysts

are much

harder

metal atoms

pattern

are participating

Ni and Co by themselves 261. As seen in Table significantly assumed

stable with

higher

rules The absence

atoms

not affect

we observed

are in the Moo2 the overall

activity.

atom does make a difference. temperatures

than

side Fe and Al containing have lower activities.

Perhaps

with molybdenum.

in converting

are quite

by X-ray

crystallites.

higher)

The fact

(less than 20 :) or atomic

CO2 feed gives significant

active

3 the activities

evidence

CO2 to methane

in converting

of the heterometal

that the hetero-

and higher

hydrocarbons.

CO2 to hydrocarbons

of Ni and Co containing

than the AHM based catalysts.

that a small fraction

denum oxide

and therefore

oxides

atoms

at lower reduction

catalysts

The activity

have been

of magnitude

The low activities

If the heterometal

to activate

should

individual

have been observed.

large heterometal

reduced

to their

(more than two orders

of the heterometal

separate

not be easily

these atoms form highly

our cata-

in the Moo2 matrix.

the presence

is some evidence

Ni and Co containing

It is apparent

may have some amorphous

seem to favor the latter explanation.

However

at 573 K

their Mo2C catalyst

et al, our catalysts

Ni, Co and Fe should

of having

of the metal

reaction

even with the CO2 feed.

of X-ray peaks can be due to very small crystallites dispersion

of

at 573 K. However,

high C2' selectivity

activities

Higher

that we could not detect

tested

of these catalysts,

atoms to molybdenum

phase

heteroatoms

out the possibility

higher temperature the C2+ selectivity

atoms

The ratio of the heterometal heteroatoms

all their bulk cata-

to as high as 30%. Again,

C2+ selectivity

MO species

Since

catalysts,

if they were

C2+ selectivity

of relatively

in the work

lower valence

higher

C2+ selectivity

not bring their

As indicated

C2' selectivity.

for molybdenum

be somewhat

would

at 773 K while

This

catalysts

is explicable

[15,20, are

if it is

atoms are on the surface

of molyb-

crystallites.

CONCLUSIONS We have found that the decomposition moderately cannot

active

Fischer-Tropsch

be detected

While

composition

is significant,

mined

product

synthesis

by X-ray analysis,

in the Moo2 matrix.

products

the effect

of 6-heteropolymolybdates

catalysts.

indicating

that

of the heteroatom the selectivity

by the Moo2 phase and does not change

The presence

of the heteroatom

it is uniformly

dispersed

on the activity

appeared

are

of the de-

to be largely

much for all the catalysts.

deter-

72

We also found that harder

to reduce

sinter

less. Feeding

in general

(when compared with

the charcoal

CO2 instead

but the C2+ selectivity

catalysts,

supported

to the unsupported

of CD decreased

is relatively

catalysts

ones),

are somewhat

but are more active

the activity

and

of most of our

high at 30 wt%.

ACKNOWLEDGEMENTS Partial Research

support

of this research

is gratefully

by the University

of Michigan

Office

of Energy

acknowledged.

REFERENCES "Third International Conference on the Chemistry and Uses of Molybdenum", Climax Molybdenum Company, August (1979) Session III, IV, V-A and VI. "Fourth International Conference on the Chemistry and Uses of Molybdenum", Climax Molybdenum Company, August (1982) 343. C.B. Murchison, "Fourth International Conference on the Chemistry and Uses of Molybdenum", Climax Molybdenum Company, August (1982) 197. C.B.Murchison and D.A. Murdick, "Use Syngas for Olefin Feedstock", Hydrocarbon Processing, January (1981) 159. J.W. Dun, E. Gulari and S. Ng, "Fischer-Tropsch Synthesis on Charcoal-supported Molybdenum; the Effect of Preparation Conditions and Potassium Promotion on Activity and Selectivity", Applied Catalysis, 15 (1985) 247. G.A. Tsigdinos, "Heteropoly Compounds of Molybdenum and Tungsten", Topics in Chemistry, Vol. 76 (1978). M. Saito and R.B. Anderson, J. Catalysis, 63 (1980) 438. J.F. Schultz, F.S. Karn and R.B. Anderson, U.S. Bureau of Mines, Report of Investigation, (1967) 6974. I. Kojima, E. Miyazaki and I. Yasumori, J. Chem. Sot., Chem. Commun., (1980) 573. A.J. Bridgewater, R. Burch and P.C.H. Mitchell, J. Catalysis, 78 (1982) 116. P.H. Emmett (Ed.), "Catalysis," Vol. 4, Pages 257, 299 and 473. Reinhold, New York (1951). J.M. Dew, R.R. White and C.M. Sliepcevitch, Ind. Eng. Chem., 47 (1955) 140. Academic Press, New York, (1962). G. Bond, "Catalysis by Metals", p. 369, J.F. Schultz, F.S. Karn and R.B. Anderson, U.S. Bureau of Mines, Report of Investigation (1967) 6941. M. Saito and R.B. Anderson, J. Catalysis, 67 (1981) 296. G.D. Weatherbee and C.H. Bartholomew, J. Catalysis, 77 (1982) 460. J. Muller, V. Pour and A. Regner, J. Catalysis, 11 (1968) 326. V.M. Vlasenko and G.E. Yuzefovich, Lisp. Khim., 38 (1969) 1622. G.A. Mills and F.W. Steffgen, Catal. Rev., 8 (1973) 159. P.J. Lunde and F.L. Kester, J. Catalysis, 30 (1973) 423. J.M. Cognion and J. Margeren, Kinet. Katal., 16 (1975) 1552. M. Araki and V. Ponec, J. Catalysis, 44 (1976) 439. R.B. Anderson, C.B. Lee and J.C. Machiels, Canad. J. Chem. Eng., 54 (1976) 590. D.J. Dwyer and G.A. Somorjai, J. Catalysis, 52 (1978) 291. J.L. Falconer and A.E. Zagli, J. Catalysis, 62 (1980) 280. R. Maatman and S. Hiemstra, J. Catalysis, 62 (1980) 349.