Catalytic properties of H2-reduced MoO3 with noble metal for the conversions of heptane and propan-2-ol

Article abstract of DOI:10.1039/a908794j

Effects of H₂ reduction on the catalytic properties of MoO3 with Pt, Pd, Rh, Ir, or Ru for the conversions of heptane and propan-2-ol were studied. The catalytic activity of MoO₃ with noble metal for the isomerization of heptane was strongly dependent on the period of H₂ reduction at 623 K. This behavior was almost the same as that of MoO₃ without noble metal. Among the catalysts tested, Pt/MoO₃ was the most active for this reaction. The catalytic activities of MoO₃ for dehydration and dehydrogenation of propan-2-o1 also increased in proportion with the period of H₂ reduction. In the case of MoO₃ with noble metal, a higher dehydration activity was obtained by a longer period of H₂ reduction, while the dehydrogenation activity was almost independent of the reduction period. Pt/MoO₃ exhibited a high dehydration activity compared with the other catalysts, indicating the most acidic property of Pt/MoO₃. We conclude from these results that the high isomerization activity of Pt/MoO₃ can be attributed to its high acidity as well as to the hydrogenative and dehydrogenative properties of Pt metal.

Full text of DOI:10.1039/a908794j

View Article Online / Journal Homepage / Table of Contents for this issue
Catalytic properties of H -reduced MoO with noble metal for the
2
3
conversions of heptane and propan-2-ol
Fumiko Uchijima, Tomoaki Takagi, Hidenobu Itoh, Takeshi Matsuda* and Nobuo Takahashi
Department of Materials Science, Kitami Institute of T echnology, Koen-cho 165, Kitami,
Hokkaido 090-8507, Japan. E-mail: MAT SUDA-T akeshi/mtrl=king.cc.kitami-it.ac.jp;
Fax: ]81 157 26 4973; T el: ]81 157 26 9448
Received 4th November 1999, Accepted 6th January 2000
E†ects of H reduction on the catalytic properties of MoO with Pt, Pd, Rh, Ir, or Ru for the conversions of
2
3
heptane and propan-2-ol were studied. The catalytic activity of MoO with noble metal for the isomerization
3
of heptane was strongly dependent on the period of H reduction at 623 K. This behavior was almost the same
2
as that of MoO without noble metal. Among the catalysts tested, Pt/MoO was the most active for this
reaction. The catalytic activities of MoO for dehydration and dehydrogenation of propan-2-ol also increased
in proportion with the period of H reduction. In the case of MoO with noble metal, a higher dehydration
activity was obtained by a longer period of H reduction, while the dehydrogenation activity was almost
independent of the reduction period. Pt/MoO exhibited a high dehydration activity compared with the other
catalysts, indicating the most acidic property of Pt/MoO . We conclude from these results that the high
isomerization activity of Pt/MoO can be attributed to its high acidity as well as to the hydrogenative and
dehydrogenative properties of Pt metal.
3
3
3
2
3
2
3
3
3
di†erence in the product distributions between these two cata-
lysts. Both dehydration and dehydrogenation of propan-2-ol
were catalyzed by the H -reduced MoO , which was more
Introduction
2
3
Molybdenum is an important catalytic component, and Mo
sulÐde- and oxide-based materials have practically been
adopted as industrial catalysts for hydrodesulfurization of pet-
roleum feed stocks and selective oxidation of alkenes, respec-
tively. A large number of research groups have extensively
investigated the catalytic properties of these materials in rela-
tion to their surface properties so as to develop more efficient
catalysts. Recently, Katrib and co-workers1,2 reported a new
catalytic function of molybdenum oxide: the MoO phase
was able to catalyze the skeletal isomerization of alkane with
a high selectivity. They proposed, from the results of system-
atic studies using catalytic reactions in association with spec-
troscopic techniques, that the isomerization reaction on
active for the dehydration than USY. We suggested on the
basis of these experimental results that the H -reduced MoO
2
3
possessed a bifunctional property, which was responsible for a
high isomerization activity.8,9
The H -reduced MoO with high catalytic activities for the
2
3
conversions of heptane and propan-2-ol has a surface area of
about 180 m2 g~1 and its average oxidation number of Mo is
in the range 3.5 to 2.5. When reduced at 623 K, a longer
reduction period is required to attain a higher catalytic activ-
ity. It has been reported by Bond and Tripathi10 that
2
reduction of MoO by molecular hydrogen proceeded only
above 673 K, but if hydrogen atoms were supplied or created
3
at the surface of MoO , reduction took place at a much lower
3
MoO proceeded via a bifunctional mechanism. Dehydroge-
temperature. This phenomenon is interpreted in terms of
2
nation and hydrogenation reactions might occur on the metal-
hydrogen spillover. We showed in a previous paper11 that
lic sites represented by a certain density of state of the free
valence electrons localized along the C-axis of a rutile
deformed lattice. Electronegative oxygen might act as an acid
site by Ðxing dissociated hydrogen produced on the metallic
sites, thus providing BrÔnsted acid sites.
loading of Pd onto MoO was e†ective in enhancing the
3
isomerization activity. The isomerization activity of Pd/MoO
3
was more strongly dependent on the period of H reduction
2
than that of MoO , although the reducibility of MoO was
3
3
markedly improved in the presence of Pd. These results indi-
Carbon-modiÐed MoO and oxygen-modiÐed Mo C were
cate that the enhanced isomerization activity of Pd/MoO can
not be explained only by the hydrogenative and dehydroge-
3
2
3
also reported to be active for the isomerization of hexane and
heptane.3h6 These Mo-based materials became active for the
isomerization after a long induction period. Since the modiÐed
MoO and Mo C exhibited di†erent catalytic behaviors from
native properties of Pd metal. The presence of Pd was found
to a†ect the surface area of H -reduced MoO as well as the
2
3
isomerization activity.
Physical and catalytic properties of H -reduced MoO with
3
2
a typical bifunctional catalyst, Pt/zeolite, the formation of a
new catalytically active phase, molybdenum oxicarbide and a
bond shift mechanism via a metallocyclobutane intermediate
were proposed by Ledoux and co-workers.4,5
2
3
noble metal are thus considered to be dependent on the kind
of noble metal, since noble metals possess di†erent abilities for
H
activation. The aim of the present study is to investigate
2
It was reported in a previous work7 that MoO became
the e†ect of H reduction on the catalytic properties of MoO
3
2
3
an active and selective catalyst for the isomerization of
with noble metal for the conversion of heptane. Dehydration
and dehydrogenation of propan-2-ol were also performed to
describe the e†ect of noble metal loading on the bifunctional
property of H -reduced MoO .
heptane after H reduction at 623 K. The isomerization activ-
2
ity of the H -reduced MoO was high compared with that of
2
3
0.5 wt.% Pt/USY zeolite, although there was no appreciable
2
3
DOI: 10.1039/a908794j
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083
1077
This journal is ( The Owner Societies 2000

View Article Online
during complete reoxidation to MoO . Since the amount of
Experimental
3
H O formation was found to be sensitive to treatment tem-
2
perature, reoxidation was performed at 523 K and then at 773
Materials
K using pulse techniques. The reduced MoO was cooled to
3
Heptane and propan-2-ol, of purity 99.3 and 99.5%, respec-
523 K in a H atmosphere, and then was kept for 1 h in a
2
tively, were dried using a molecular sieve prior to use. H , He
stream of He, followed by injection of O . After no consump-
tion of O was observed at this temperature, the sample was
heated to 773 K in a He atmosphere, and then O was
injected again. The average oxidation number of Mo was cal-
2
and N gases were puriÐed by passing through a molecular
2
2
2
sieve and a Mn/SiO oxygen trap. H MoO , of purity 98%,
2
2
4
2
was purchased from Kanto Chemical Co., Inc. Commercially
available noble metal chloride, PtCl É 5H O, PdCl ,
culated from the amount of O consumption at 523 K plus
4
2
2
2
RhCl É 3H O, RuCl É nH O and IrCl É nH O, were used
that at 773 K. H O was formed in heating from 523 to 773 K
3
2
3
2
4
2
2
without further puriÐcation.
as well as in the reoxidation processes. The total amounts of
H O formation are shown in this work.
2
Preparation of catalyst
Acidity measurement
The MoO used in this study was obtained by calcination of
3
H MoO at 673 K for 3 h. MoO with 0.01 mol% noble
Acidity of MoO with and without noble metal was deter-
2
4
3
metal was prepared by a conventional impregnation method
3
mined using the NH ÈTPD technique. In each NH ÈTPD
using an aqueous solution of noble metal chloride. MoO (20
3
3
experiment, a sample weighing 0.1 g was placed in a cell, and
then was reduced at 623 K for 48 h. After cooling to 373 K
and replacement of H by He, NH was adsorbed at this tem-
perature for 0.5 h. A stream of He was passed through the
apparatus to displace NH for 1 h, and then the sample was
heated to 1023 K at a rate of 10 K min~1. A thermal conduc-
tivity detector was used to monitor desorbed NH . Since
water was desorbed simultaneously with NH , the NH ÈTPD
spectrum was obtained by point-by-point subtraction of the
water desorption spectrum of the sample, which had not
3
g) was added to 20 ml of 7 ] 10~4 M solution, and then
evaporation of the solution was carried out at 333 K. The
obtained sample was dried overnight at 393 K, followed by
calcination at 673 K for 3 h. The catalyst powder was made
into pellets, crushed, and then sieved (30È60 mesh) for charg-
ing into a reactor.
2
3
3
3
Catalytic test
3
3
Reactions of heptane and propan-2-ol were isothermally
carried out at 523 and 398 K, respectively, under atmospheric
pressure in a conventional Ðxed bed Ñow reactor equipped
with a gas sampling valve for gas chromatographic analysis. A
prescribed amount of catalyst (0.05È0.2 g) was packed at the
central position of a reactor. It was then heated to 623 K at a
adsorbed NH .
3
Results
Conversion of heptane
rate of 5 K min~1 in a stream of H and kept for a desired
2
period. It was shown in a previous paper12 that the physical
H -reduced MoO catalysts exhibited almost stable activities
2
3
and catalytic properties of reduced MoO were signiÐcantly
against deactivation in the conversion of heptane. Here, the
catalytic activity was estimated using data after a 6 h run. Fig.
3
inÑuenced by the Ñow rate of H in the reduction process.
2
Hence, reduction was performed at the H Ñow rate of 600 ml
min~1 g-MoO ~1 in this study. After cooling to reaction tem-
perature under H atmosphere, heptane or propan-2-ol was
1 shows the catalytic activities of reduced MoO with and
without noble metal as a function of reduction period.
2
3
3
Without H reduction, all of the catalysts were virtually inac-
2
2
introduced onto the catalyst bed at the pressure of 2471 or
tive for the conversion of heptane. H reduction at 623 K
2
4825 Pa with H as a complement to atmospheric pressure.
markedly improved the catalytic activity. In the cases of
2
For comparison, reaction of propan-2-ol was also carried out
MoO , Rh/MoO , Ir/MoO and Ru/MoO , the catalytic
3
3
3
3
using N as a carrier gas. The composition of effluent gases
activities became almost constant after 12 h reduction. The
activities of Pt/MoO and Pd/MoO , however, were enlarged
2
was analyzed by means of FID gas chromatography using a
3
3
TC-1 glass capillary separation column, and TCD gas chro-
matography using a unibead 1S separation column.
by a further period of reduction, and the constant activities
were obtained after reduction for 48 h. Pt/MoO was the most
active for the conversion of heptane among the catalysts
tested.
3
Characterization methods
Crystalline phases of the catalyst were determined by X-ray
di†raction using Ni-Ðltered Cu-Ka radiation (Rigaku, Rint-
1200). The sample for XRD measurements was obtained as
follows: MoO with and without noble metal was subjected
3
to H reduction at 623 K for 48 h, followed by Ñowing N for
2
2
0.5 h at the same temperature. After cooling to room tem-
perature under N atmosphere, the sample was passivated for
2
1 h with 1% O È99% N to avoid a deep bulk oxidation.
2
2
Surface area was determined from a N adsorption iso-
therm, which was obtained on the sample without exposure to
2
air. The reduced sample was cooled to room temperature with
H still Ñowing, and then was evacuated for 0.5 h, followed by
2
the measurement of N adsorption at 77 K using a conven-
2
tional high vacuum static system. To determine pore size dis-
tribution, adsorptionÈdesorption isotherms of
N
were
2
measured with an automatic gas adsorption apparatus
(Sorptomatic 1990, Carlo Erba) after evacuation at 573 K for
2 h.
Fig. 1 Variation in % conversion of heptane as a function of H
2
reduction period on MoO (…), Ru/MoO (>), Ir/MoO (È),
3
3
3
The average oxidation state of Mo and the quantity of H
retained in the reduced sample were determined by measuring
Rh/MoO (K), Pd/MoO (|) and Pt/MoO (L). Reaction condi-
2
3
3
3
tions: temperature, 523 K; H /C , 40; weight of catalyst/Ñow rate of
2
7
the amount of O consumption and that of H O formation
C , 25 g-cat h mol-C ~1. Data were taken after a 6 h run.
7
7
2
2
1078
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083

View Article Online
Under the reaction conditions employed, heptane was selec-
tively isomerized with a minor contribution of cracking on all
of the catalysts. Fig. 2 demonstrates the isomerization selec-
tivity of the catalysts reduced for 48 h vs. the conversion level.
There was no appreciable di†erence in the isomerization selec-
tivity among the catalysts tested, indicating that noble metal
expense of C and C hydrocarbons. Hydrocracking of
3
4
heptane seems to be catalyzed by noble metal on MoO .
3
Conversion of propan-2-ol
Conversion of propan-2-ol was performed at 398 K in a H
2
atmosphere. The catalytic activity of H -reduced MoO in
loading onto MoO gave little e†ect on the reaction selec-
2
3
3
this reaction declined gradually with time on stream, regard-
less of the reduction period. Catalyst deactivation was hardly
a†ected by loading of noble metal. The catalytic activities of
tivity. An increase in the conversion level was accompanied by
a decrease in the isomerization selectivity. High isomerization
selectivity, however, was attained even at high conversion
levels.
MoO with and without noble metal for the conversion of
3
propan-2-ol were compared using data after a 6 h run in the
Table 1 summarizes the product distributions on the cata-
lysts reduced for 48 h. Here, the conversion level was adjusted
to be about 40% by varying Ñow rate of reactant gas. No
appreciable di†erence was observed in the distribution of
isomerization products among the catalysts tested. Heptane
was mainly isomerized to 2- and 3-methylhexanes (2-MH and
3-MH) in equal amounts. At high conversion levels, dimethyl-
pentanes (DMP) were formed at the expense of methyl-
hexanes. By contrast, the distribution of cracking products
was altered by loading of noble metal. The cracking reaction
on the reduced MoO mainly yielded C and C hydrocar-
following section.
Under the reaction conditions employed, propan-2-ol was
converted to propene and diisopropylether (DIPE) by dehy-
dration and to acetone by dehydrogenation. Fig. 3 shows the
variation in formation rate of each product on the reduced
MoO catalyst with the reduction period. Here, the formation
3
rates were determined using data at the conversion level
below 10%. The catalytic activities of MoO for dehydration
3
and dehydrogenation were enlarged by
H
reduction,
2
although the dehydrogenation activity was markedly low
compared with the dehydration activity even after 48 h
3
3
4
bons. The distribution of cracking products on this catalyst
was accounted for by single CÈC bond breaking. In the case
of the reduced MoO with noble metal, the formation of C
reduction. The reduced MoO produced propene much more
3
selectively than DIPE. Results of Pt/MoO are demonstrated
3
3
1
in Fig. 4. The dehydration activity of Pt/MoO increased in
and C hydrocarbons, especially C , was enlarged at the
3
6
1
Fig. 3 Catalytic activity of MoO for the conversion of propan-2-ol
3
Fig. 2 Dependency of the isomerization selectivity on the conversion
level of heptane. Catalyst: …, MoO ; >, Ru/MoO ; È, Ir/MoO
as a function of H reduction period. L, C ; |, DIPE; K, acetone.
2
2
;
Reaction conditions: temperature, 398 K; H /propan-2-ol, 20. Data
3
3
3
2
K, Rh/MoO ; |, Pd/MoO ; L, Pt/MoO .
were taken after a 6 h run.
3
3
3
Table 1 Product distributions on MoO with and without noble metal in the conversion of heptanea
3
Catalyst
MoO
35.6
Pt/MoO
42.6
Pd/MoO
39.5
Rh/MoO
43.1
Ir/MoO
37.2
Ru/MoO
45.7
3
3
3
3
3
3
% Conversion
Selectivity (C-mol.%)
Isomerization
Cracking
96.8
2.6
97.3
2.1
96.8
1.9
97.5
1.8
96.5
2.2
96.6
2.3
Distribution (mol.%)
Isomerization
2-MH
42.1
40.5
14.7
2.7
39.1
42.7
13.7
4.5
40.2
44.5
13.1
2.2
40.2
43.9
13.6
2.3
39.6
45.3
13.1
2.0
40.1
44.7
13.3
1.9
3-MH
DMP
Others
Cracking
C
4.2
4.0
42.2
43.5
3.7
18.4
4.9
28.6
37.6
4.2
25.2
6.9
21.7
31.3
5.7
21.9
5.9
24.7
34.4
5.0
23.5
7.7
22.4
31.4
6.2
25.6
8.1
21.2
28.4
6.7
1
C
2
C
3
C
4
C
5
C
6
2.4
6.3
9.2
8.1
8.7
10.0
a Catalysts were reduced at 623 K for 48 h.
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083
1079

View Article Online
of MoO . Loading of Pt, Pd, Rh, and Ir onto MoO ,
3
3
however, was found to improve the dehydration activity.
Among the catalysts tested, Pt/MoO was most active both
3
for the dehydration and the dehydrogenation.
Characteristics
It was shown in a previous work13 that H reduction of
2
MoO at 623 K was accompanied by an increase in the spe-
3
ciÐc surface area, and yielded a porous material, of which the
pore radius was in the range from 0.3 to 1.5 nm. Hence, char-
acteristics of the noble metal-loaded MoO catalysts, which
3
were reduced for 48 h, were studied. Fig. 6 illustrates the XRD
patterns of the reduced MoO . The di†raction lines corre-
3
sponding to MoO was not observed at all on all of the cata-
3
lysts. In the cases of Pt, Pd, Rh, and Ir/MoO , the di†raction
lines were observed only at 2h \ 38.1, 40.7, and 44.3¡. The line
at 40.7¡ is assigned to d(110) plane of Mo metal. Hydrogen
3
Fig. 4 Catalytic activity of Pt/MoO for the conversion of propan-2-
3
ol as a function of H reduction period. L, C ; |, DIPE; K, acetone.
2
2
Reaction conditions: temperature, 398 K; H /propan-2-ol, 20. Data
molybdenum bronze, H MoO , provides the di†raction line
2
were taken after a 6 h run.
2
3
of d(600) at 40.0¡. This line appears at a lower angle when the
content of decreases: the d(600) di†raction line of
H
2
H
MoO is observed at 2h \ 37.2¡. Thus, the di†raction
proportion with the period of reduction. This behavior is
0.93
3
line appearing at 2h \ 38.1¡ seems to be related to hydrogen
molybdenum bronze, H MoO . Delporte and co-workers6
reported that a di†raction line of molybdemun oxicarbide,
almost identical with that of MoO . The dehydrogenation
3
activity of Pt/MoO , however, was independent of the
x
3
3
reduction period. Similar results were obtained on MoO with
other noble metals. Reaction of propan-2-ol was carried out
3
MoO C , was observed at 2h \ 44.3¡. They stated that H or
x y
H
was able to act as C atoms to form MoO H , of which a
using N as a carrier gas to study the e†ect of hydrogen spill-
2
x y
2
di†raction line appeared at 2h \ 44.5¡. The line at 2h \ 44.3¡
over on the catalytic activities of H -reduced MoO . Fig. 5
2
3
seems to reÑect the formation of MoO H . The reduced
shows the results of Pt/MoO . The dehydration activity of
x
y
3
MoO and Ru/MoO gave the di†raction lines corresponding
to MoO phase at 2h \ 26.0, 36.9 and 53.7¡ as well as the
above three lines.
Pt/MoO in N atmosphere was comparable with that in H .
3
3
3
2
2
Table 2 summarizes the dehydration and dehydrogenation
2
activities of MoO catalysts reduced for 48 h. Noble metal-
3
The average valence number of Mo was calculated from the
loaded MoO exhibited a higher dehydrogenation activity
3
amount of O consumption in complete reoxidation to
MoO . As shown in Table 3, average valence number of Mo
in the reduced MoO was determined to be 2.7. The reduced
Ru and Ir/MoO exhibited almost the same valence number
as the MoO . The average valence number of Mo in the
reduced Pt, Pd, and Rh/MoO was lower than that of MoO ,
although the di†erence was small. Reoxidation of the reduced
MoO catalysts was accompanied by the formation of H O,
suggesting the presence of H in the reduced samples. This
than MoO , due to the promotive e†ect of noble metal. The
dehydration activity of Ru/MoO was comparable with that
2
3
3
3
3
3
3
3
3
3
2
2
x
may reÑect the formation of H MoO and/or MoO H . There
was no appreciable di†erence in the amount of H retained
among the catalysts tested.
3
x y
2
The unreduced MoO , irrespective of noble metal loading,
3
had a BET surface area of 5 m2 g~1. For the reduced MoO
3
catalysts, however, the Langmuir isotherm provided a satisfac-
Fig. 5 Catalytic activity of Pt/MoO for the conversion of propan-2-
3
ol in N atmosphere as a function of H reduction period. L, C ; |,
2
2
2
DIPE; K, acetone. Reaction conditions: temperature, 398 K;
N /propan-2-ol, 20. Data were taken after a 6 h run.
2
Table 2 Catalytic activity of H reduced MoO with and without
2
3
noble metal for the conversion of propan-2-ola
Formation rate/10~3 mol h~1 g~1
Catalyst
C
DIPE
Acetone
3
MoO
3
Pd/MoO
Rh/MoO
Ir/MoO
16.87
37.25
22.67
22.83
20.94
16.35
0.35
0.57
0.18
0.26
0.31
0.33
0.19
2.62
0.73
0.52
1.24
0.45
Pt/MoO
3
3
3
3
Ru/MoO
3
Fig. 6 XRD patterns of MoO (A), Ru/MoO (B), Ir/MoO (C),
3
3
3
Rh/MoO (D), Pd/MoO (E) and Pt/MoO (F). The catalysts were
reduced for 48 h.
a Catalysts were reduced with H at 623 K for 48 h.
3
3
3
2
1080
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083

View Article Online
Table 3 Physical properties of H reduced MoO with and without
curves were obtained by analysis of N adsorption data using
the HorvathÈKawazoe method. Typical results are illustrated
2
3
2
noble metala
in Fig. 7. The reduced MoO had micropores, of which the
Average
oxidation
state of Mo
Amount of
retained
/10~4 mol g~1
3
pore radius was in the range 0.3 to 1.5 nm. Pt and Ru/MoO
Surface area
/m2 g~1
H
3
had the same pore size distributions as MoO , indicating that
3
2
Catalyst
the formation of micropores was hardly a†ected by loading of
MoO
Pt/MoO
Pd/MoO
Rh/MoO
180
216
208
230
218
229
2.7
2.3
2.1
2.2
2.6
2.9
4.4
4.4
3.6
3.3
4.5
3.9
noble metal.
3
3
3
Acidity
3
Ir/MoO
Acidity of H -reduced MoO catalysts was measured by
means of the NH ÈTPD technique. Fig. 8 demonstrates the
results of MoO , Pt/MoO and Ru/MoO . The NH ÈTPD
spectrum of reduced MoO consisted of two desorption peak
3
Ru/MoO
3
2
3
3
a Catalysts were reduced with H at 623 K for 48 h.
2
3
3
3
3
3
maxima; the low and high temperature maxima appeared at
about 460 and 773 K, respectively. Pt and Ru/MoO provided
tory linear Ðt of the N adsorption data, while the BET did
not. Hence, speciÐc surface areas were calculated using the
Langmuir equation. Results are demonstrated in Table 3.
3
2
a similar desorption proÐle of NH to MoO . As shown in
3
3
Table 4, the amounts of desorbed NH from noble metal-
3
loaded MoO were greater than that from MoO , indepen-
MoO reduced for 48 h had a surface area of 180 m2 g~1.
3
3
3
dent of the kind of noble metal. When compared with the
density of acid sites, however, no appreciable di†erence was
observed, indicating that noble metal loading induced the gen-
MoO with noble metal exhibited a larger surface area than
3
MoO after H reduction for 48 h. Their surface areas were in
3
2
the range 200 to 230 m2 g~1. The reduced samples with a
large surface area provided N adsorptionÈdesorption iso-
therms with no hysteresis and the amount of N adsorption
increased sharply at very low pressures and changed little at
eration of a large number of acid sites on MoO due to the
3
2
enlargement of surface area.
2
Discussion
intermediate pressures. These phenomena are characteristics
of microporous materials. Therefore, pore size distribution
Surface area of H -reduced samples
2
H
reduction at 623 K signiÐcantly enlarges surface areas of
2
MoO with and without noble metal. This phenomenon has
already been observed in temperature-programmed reaction
3
with NH ,14,15 and in carburization using a H and CH
3
2
4
mixture.16 H reduction of MoO at 623 K has also been
2
3
reported to produce a mixture of MoO and MoO H phases,
2
x y
of which the surface area was about 150 m2 g~1.6 It was
shown in our previous work13 that H -reduced MoO with a
2
3
large surface area gave the di†raction line at 2h \ 44.3¡ which
reÑected the formation of the MoO H phase, while that with
x
y
a low surface area did not. As shown in Fig. 6, the H -reduced
2
MoO with a large surface area gives the di†raction line at
3
2h \ 44.3¡, irrespective of noble metal loading. We suggest
from these results that the enlargement of surface area can be
related to the formation of the MoO H phase. Since MoO
x
y
3
with noble metal exhibits a greater surface area than MoO
3
after H reduction, the formation and/or the stability of the
2
MoO H phase seems to be inÑuenced in the presence of
Fig. 7 Pore size distribution curves of MoO (A), Ru/MoO (B), and
x
y
3
3
Pt/MoO (C). The catalysts were reduced for 48 h.
noble metal. At present, however, we do not have any plaus-
3
ible model to explain the large surface area of H -reduced
2
samples satisfactorily. The detailed reduction process to yield
porous MoO with a large surface area is still under investiga-
x
tion.
Catalytic activity of H -reduced samples
2
Alkane isomerization has been proposed to proceed via a
bifunctional mechanism, C -ring hydrogenolysis process, or
5
metallocyclobutane bond-shift mechanism.17,18 Ledoux and
co-workers3h5 reported that the oxygen-modiÐed Mo C and
2
MoO exhibited a long induction period in the isomerization
3
of heptane at 623 K. They suggested the formation of a new
catalytically active phase, MoO C in the course of the reac-
x y
tion, and a bond shift mechanism via a metallocyclobutane
intermediate. On these catalysts, C and C hydrocarbons
1
6
were mainly produced by cracking of heptane. It was shown
in our previous paper9 that catalytic behaviors of the H -
2
reduced MoO were di†erent from those of the MoO C : no
3
x y
induction period was observed on the H -reduced MoO , and
the major cracking products were C and C . The H -reduced
MoO with noble metal exhibited almost the same behavior
2
3
3
4
2
Fig. 8 NH ÈTPD spectra of MoO (A), Ru/MoO (B), and
3
3
3
3
Pt/MoO (C). The catalysts were reduced for 48 h.
as that without noble metal.
3
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083
1081

View Article Online
Dehydration and dehydrogenation of propan-2-ol proceed
simultaneously on H -reduced MoO , irrespective of noble
the result of Ir/MoO , the order of the isomerization activity
was in accordance with that of the dehydrogenation activity.
We suggest from these results that the high isomerization
3
2
3
metal loading. The catalytic activities of MoO for both dehy-
3
dration and dehydrogenation increase in proportion with the
activities of Pd, Rh, Ir, and Ru/MoO result from the pro-
3
period of H reduction. In the case of noble metal-loaded
motion of the dehydrogenationÈhydrogenation step by the
2
MoO , the dehydrogenation activity is independent of the
noble metal. Since Pt/MoO is more active for the dehydra-
3
3
reduction period, while the higher dehydration activity is
tion than MoO , the high isomerization activity of Pt/MoO
3
3
obtained after a longer period of reduction. We suppose on
would be attributed to its high acidity as well as to the hydro-
genative and dehydrogenative properties of Pt metal.
the basis of these experimental results that H reduction can
2
improve the isomerization activity of MoO with noble metal,
3
Pt/MoO exhibits the highest dehydration activity per
3
mainly due to the generation and/or the enhancement of
surface area, although no appreciable di†erence is observed in
acidity. In contrast, the increased isomerization activity of
MoO by H reduction seems to be related to the formation
of dual sites: sites with metallic character and sites with acidic
property.
the density of acid sites and the acid strength among the cata-
lysts tested. Hydrogen spillover has been known to a†ect acid-
catalyzed reactions. In most of the cases reported,
acid-catalyzed reactions were retarded in the presence of spilt-
over hydrogen.26h31 In contrast, the promoting e†ect of spilt-
over hydrogen was observed on Ag-exchanged Y-zeolites32
and on Pt/SO 2~ÈZrO .33 Spiltover hydrogen was reported
3
2
Correlation between the oleÐn hydrogenation activity and
distribution of Mo oxidation state in the reduced MoO on
3
Al O supports was investigated in detail.19h22 Hercules and
2
3
4
2
co-workers21,22 suggested that Mo metal and Mo2` were
only the active center for benzene hydrogenation. The active
sites for hydrogenation are ordinarily capable of catalyzing
to form protonic acid sites on Pt/SO 2~ÈZrO , and conse-
4
2
quently to enhance the catalytic activities. In the case of
Pt/MoO , the dehydration activity of Pt/MoO in H atmo-
3
3
2
the dehydrogenation reaction. In the case of MoO , the for-
sphere is comparable with that in N , indicating the negligible
3
2
mation of lower oxidation state of Mo seems to promote the
e†ect of spiltover hydrogen. It is generally recognized that
dehydrogenation reaction. On the other hand, noble metal as
well as the lower oxidation state of Mo would be responsible
for dehydrogenation reaction on MoO with noble metal.
BrÔnsted acid sites are more active than Lewis ones. From the
result of dehydration, Pt/MoO is considered to be more
3
BrÔnsted acidic than other catalysts.
3
The acidity of MoO supported on Al O was widely inves-
H retention in the reduced MoO /Al O has been thor-
3
2
3
2
3
2 3
tigated. It was reported23,24 that H reduction induced a
oughly investigated by Hall and co-workers.34,35 They
2
decrease in BrÔnsted acidity of MoO /Al O . In contrast,
showed that retained H was held as MoÈOÈH and MoÈH,
which released H as H O in the reoxidation process. As
shown in Table 3, the large amounts of H are retained in all
of the reduced MoO catalysts. The amount of H , however,
is hardly a†ected by noble metal loading. H exists both in
H MoO and MoO H phases. H in molybdenum bronze,
H MoO , is unable to act as an acid site due to its neutral
3
2
3
2
2
Suarez and co-workers25 indicated that the number of
BrÔnsted acid sites on MoO /Al O was enlarged by
3
2
3
2
reduction up to an average Mo oxidation state of 5.6 and
decreased to zero with further reduction. The results of dehy-
dration obtained in this study, however, are inconsistent with
the above reported results. H -reduced MoO catalysts with
3
2
x
x
3
3
x y
2
3
average Mo oxidation states of 2.9 to 2.1 are active for the
dehydration. In dehydration of alcohol, oleÐn is preferentially
formed on strong acid sites, while weak acid sites are capable
of catalyzing ether formation. As shown in Table 1, propan-2-
ol is more selectively converted to propene on all of the cata-
lysts tested, indicating the presence of strong acid sites. This is
conÐrmed from the results of NH ÈTPD: the H -reduced
property. A part of H in the MoO H phase seems to play an
x
y
important role in generating BrÔnsted acid sites. We deduce
from the results of dehydration that the formation of the
MoO H and/or its ability to generate acid sites can be
x
y
a†ected in the presence of Pt.
Conclusions
3
2
MoO catalysts provide a desorption peak maximum at 773
3
K.
Conversions of heptane and propan-2-ol were performed
using the H -reduced MoO . The catalytic activities of MoO
To manifest the e†ect of noble metal loading, the catalytic
activity was compared on the basis of the surface area. Results
are summarized in Table 4. Here, the catalysts were reduced
for 48 h, and the activity was estimated using data at the con-
version level below 10%. The isomerization activity per
surface area decreased in the following sequence: Pt [ Pd [
Rh [ Ir [ Ru [ MoO . Pt/MoO provided the highest dehy-
2
3
3
with and without noble metal for heptane isomerization were
similarly enhanced by H reduction at 623 K. Loading of
2
noble metal onto MoO was e†ective to improve the isomer-
3
ization activity. Pt/MoO exhibited the highest isomerization
activity among the catalysts tested. The catalytic activities of
3
noble metal-loaded MoO for dehydration of propan-2-ol
3
3
3
dration activity even when the surface area was taken into
also increased in proportion with the period of H reduction,
2
consideration. The dehydration activities of MoO with other
whereas their dehydrogenation activity was hardly a†ected by
3
noble metals were comparable with that of MoO . On the
3
the reduction period. Pt/MoO exhibited high dehydration
3
other hand, the dehydrogenation activity decreased in the fol-
and dehydrogenation activities compared with other catalysts.
lowing order: Pt [ Ir [ Pd [ Rh [ Ru [ MoO . Excepting
H reduction of MoO with and without noble metal yielded
3
2
3
Table 4 Catalytic and acidic properties of the H reduced MoO with and without noble metala
2
3
Amount of desorbed NH
2
Isomerization
Dehydration
Dehydrogenation
Catalyst
/10~5 mol h~1 m~2
/10~4 mol h~1 m~2
/10~5 mol h~1 m~2
/10~4 mol g~1
/10~6 mol m~2
MoO
3
Pd/MoO
Rh/MoO
Ir/MoO
2.72
7.69
6.39
5.74
3.89
2.94
0.96
1.75
1.10
1.00
0.98
0.73
0.11
1.21
0.35
0.23
0.57
0.20
2.01
3.05
2.57
2.76
3.13
2.74
1.12
1.41
1.24
1.20
1.43
1.20
Pt/MoO
3
3
3
3
Ru/MoO
3
a Catalysts were reduced at 623 K for 48 h.
1082
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083

View Article Online
13 T. Matsuda, Y. Hirata, F. Uchijima, H. Itoh and N. Takahashi,
Bull. Chem. Soc. Jpn., in the press.
14 L. Volpe and M. Boudart, J. Solid State Chem., 1985, 59, 332.
15 J. G. Choi, R. L. Curl and L. T. Thompson, J. Catal., 1994, 146,
218.
microporous MoO with large surface areas. The results of
comparison of the catalytic activities on the basis of surface
area allow us to suggest that the high isomerization activity of
x
Pt/MoO catalyst can be attributed to its high acidity as well
3
as to the promotion of the dehydrogenationÈhydrogenation
16 J. S. Lee, S. T. Oyama and M. Boudart, J. Catal., 1987, 106, 125.
17 F. G. Gault, in Advances in Catalysis, ed. D. D. Eley, H. Pines
and P. B. Weisz, Academic Press, New York, 1981, 30, 1.
18 R. J. Anderson, R. I. MacDonald and Y. Shimoyama, J. Catal.,
1971, 20, 145.
step by Pt metal.
Acknowledgements
This work was supported in part by a Grant-in-Aid for
Encouragement of Young Scientists from the Japanese Minis-
try of Education, Science, Sports and Culture.
19 E. A. Lombardo, M. Houalla and W. K. Hall, J. Catal., 1978, 51,
256.
20 E. A. Lombardo, M. Jacono and W. K. Hall, J. Catal., 1980, 64,
150.
21 R. B. Quincy, M. Houalla, A. Proctor and D. A. Hercules, J.
Phys. Chem., 1990, 94, 1520.
22 M. Yamada, J. Yasumaru, M. Houalla and D. M. Hercules, J.
Phys. Chem., 1992, 95, 7037.
23 K. Segawa and W. K. Hall, J. Catal., 1982, 76, 133.
24 S. Rajagopal, J. A. Marzari and R. Miranda, J. Catal., 1995, 151,
192.
References
1
A. Katrib, F. Hemming, P. Wehrer, L. Hilaire and G. Maire,
Catal. L ett., 1994, 29, 397.
2
A. Katrib, V. Logie, N. Saurel, P. Wehrer, L. Hilaire and G.
Maire, Surf. Sci., 1997, 337-379, 754.
25 W. Suarez, J. A. Dumesic and C. G. Hill Jr., J. Catal., 1985, 94,
3
4
C. P. Huu, M. J. Ledoux and J. Guille, J. Catal., 1993, 143, 249.
E. A. Blekkan, C. P. Huu, M. J. Ledoux and J. Guille, Ind. Eng.
Chem. Res., 1994, 33, 1657.
408.
26 H. G. Karge, Z. Sarbak, H. Hatada, J. Weitkamp and P. A.
Jacobs, J. Catal., 1983, 82, 236.
5
M. J. Ledoux, C. P. Huu, P. Delporte, E. A. Blekkan, A. P. E.
York, E. G. Derouane and A. Fonseca, Stud. Surf. Sci. Catal.,
1994, 92, 81.
27 H. G. Karge, Y. Wada, J. Weitkamp and P. A. Jacobs, J. Catal.,
1984, 88, 251.
28 N. S. Gnep, M. L. Martinde Armando and M. Guisnet, Stud.
Surf. Sci. Catal., 1983, 17, 309.
6
7
8
9
P. Delporte, F. Meunier, C. P. Huu, P. Vennegues, M. J. Ledoux
and J. Guille, Catal. T oday, 1995, 23, 251.
29 N. S. Gnep and M. Guisnet, Appl. Catal., 1981, 1, 329.
30 T. Matsuda, T. Fuse and E. Kikuchi, J. Catal., 1987, 106, 38.
31 T. Matsuda, H. Seki and E. Kikuchi, J. Catal., 1995, 154, 41.
32 T. Baba and Y. Ono, Zeolites, 1987, 7, 292.
33 H. Hattori, K. Ebitani, T. Tanaka and K. Tanabe, in Catalytic
Science and T echnology, ed. S. Yoshida, N. Takezawa and T.
Ono, Kodansha, VCH, Weinheim, 1991, vol. 1, p. 117.
34 W. K. Hall and F. E. Massoth, J. Catal., 1974, 34, 41.
35 W. K. Hall and M. L. Jacono, in Proceedings of the 6th Interna-
tional Congress on Catalysis, London, 1977, p. 246.
T. Matsuda, H. Shiro, H. Sakagami and N. Takahashi, Catal.
L ett., 1997, 47, 99.
T. Matsuda, Y. Hirata, H. Sakagami and N. Takahashi, Chem.
L ett., 1997, 12, 1261.
T. Matsuda, Y. Hirata, S. Suga, H. Sakagami and N. Takahashi,
Appl. Catal. A, 2000, 193, 185.
10 G. C. Bond and J. B. Tripathi, J. Chem. Soc., Faraday T rans. 1,
1976, 72, 933.
11 T. Matsuda, F. Uchijima, S. Endo and N. Takahashi, Appl. Catal.
A, 1999, 176, 91.
12 T. Matsuda, Y. Hirata, M. Suzuki, H. Sakagami and N. Taka-
hashi, Chem. L ett., 1999, 9, 873.
Paper a908794j
Phys. Chem. Chem. Phys., 2000, 2, 1077È1083
1083