H2-reduction of Pt/MoO3 to MoOx with a large surface area and its catalytic activities for the conversions of heptane and propan-2-ol

Article abstract of DOI:10.1039/b105680h

The relationship between the surface area of H₂-reduced Pt/MoO₃ and the extent of reduction was studied. The surface areas of H₂-reduced MoO₃ and Pt/MoO₃ varied significantly with reduction temperature and period. The largest surface area was obtained at a reduction degree of 60-70%. The distribution of isomerization products was independent of the reduction degree. Heptane was isomerized to 2-and 3-methylhexane in equal amounts. The isomerization selectivity decreased slightly with increasing conversion level, but the high selectivity was preserved even at high conversion levels on the catalysts. The catalytic behavior of H₂-reduced MoO₃ and Pt/MoO₃ in the isomerization of heptane seemed to be different from that in the conversion of 2-propanol.

Full text of DOI:10.1039/b105680h

View Article Online / Journal Homepage / Table of Contents for this issue
H -reduction of Pt/MoO to MoO with a large surface area and its
2
3
x
catalytic activities for the conversions of heptane and propan-2-ol
Takeshi Matsuda,* Fumiko Uchijima, Hirotoshi Sakagami and Nobuo Takahashi
Department of Materials Science, Kitami Institute of T echnology, Koen-cho 165, Kitami,
Hokkaido 090-8507, Japan. E-mail: matsutk=king.cc.kitami-it.ac.jp; Fax ]81 157 26 4973
Received 28th June 2001, Accepted 14th August 2001
First published as an Advance Article on the web 11th September 2001
The surface area of H -reduced 0.01 mol.% Pt/MoO increased in proportion to the extent of reduction, and
2
3
reached the maximum value of 250 m2 g~1 at a reduction degree of 60È70%. In the case of H -reduced MoO ,
2
3
the relationship between the surface area and the reduction degree was dependent on reduction temperature.
MoO reduced at 673 K exhibited a much smaller surface area than that at 623 K even when the reduction
3
degree was comparable. XRD studies showed that reduction of Pt/MoO proceeded via the formation of
3
H MoO phase, irrespective of reduction temperature. In the reduction of MoO without Pt, however, the
H MoO phase was formed at temperature below 623 K. The heptane isomerization activity of H -reduced
x
x
3
3
3
2
MoO and Pt/MoO was similarly dependent on the reduction degree as the surface area. H -reduced MoO
3
3
2
3
and Pt/MoO catalyzed the dehydrogenation and dehydration of propan-2-ol simultaneously, indicating the
3
presence of dual sites. There were good relationships between the isomerization activity and the dehydration
activity. We suggest from these results that the surface area and the catalytic activities can be enlarged when
reduction of MoO proceeds through the formation of a H MoO phase.
3
x
3
We reported that the surface area and the catalytic activity
of H -reduced MoO for heptane isomerization were depen-
dent on the reduction degree. The relationship, however, dif-
Introduction
2
3
It is well known that Mo-based materials, such as oxides,
sulÐde and carbide, exhibit interesting catalytic properties for
several hydrocarbon conversions. Ledoux and co-workers
have reported that molybdenum oxicarbide, MoO C , cata-
lyzed the skeletal isomerization of hexane and heptane
fered from reduction temperature.13,14 The surface area and
the isomerization activity of MoO reduced at 623 K were
3
much higher than those of MoO reduced at 673 K even when
x y
3
the reduction degree was comparable. XRD studies showed
without cyclization and excessive cracking.1h3 They proposed
that reduction of MoO at 623 K proceeded via the formation
3
a bond-shift mechanism via a metallocyclobutane intermediate
of hydrogen molybdenum bronze, H MoO , while MoO was
x
3
3
on the MoO C phase. Katrib and co-workers have shown
directly reduced to MoO at 673 K. We have suggested from
x y
2
that the isomerization of 2-methylpentane was selectively
these results that the surface area and the isomerization activ-
catalyzed by the MoO phase.4,5 They interpreted the isomer-
ity can markedly be enlarged when MoO is reduced through
2
3
ization reaction on the MoO phase in terms of a bifunctional
the formation of a H MoO phase.
2
x
3
mechanism. Dehydrogenation and hydrogenation might occur
Sermon and Bond have reported that reduction of MoO
3
on the metallic sites represented by a certain density of state
of the free valence electrons localized along the C-axis of the
rutile deformed lattice. Electronegative oxygen might act as
acid sites by Ðxing dissociated hydrogen produced on the
by molecular hydrogen occurs only above 673 K, but if hydro-
gen atoms are supplied or created at the surface of MoO3
then reduction takes place at much lower temperatures.15
This phenomenon is interpreted in terms of hydrogen spill-
metallic sites, thus providing Bronsted acid sites. H -reduced
over. It has been shown that MoO with noble metal absorbs
2
3
MoO /a-Al O was also reported to catalyze the isomer-
a large amount of H without an equivalent amount of water
3
2 3
2
ization of hexanes by a bifunctional mechanism.6,7
being generated, and reduction of MoO proceeds via the for-
3
We showed in the previous papers8,9 that H reduction of
mation of the H MoO phase in the presence of spilt-over
hydrogen.16h18 Loading of a noble metal onto MoO is thus
expected to provide a positive e†ect on the enlargement of the
2
x
3
MoO was accompanied by an increase in the surface area.
3
3
3
MoO reduced at 623 K for 12 h and longer exhibited a
surface area of about 180 m2 g~1. Furthermore, the reduced
surface area and of the isomerization activity. We showed in
sample was found to have pores with diameters of 6 to 30 A
Ž
.
previous papers that MoO with noble metal exhibited a
3
Our previous papers clariÐed that MoO was converted into
an active and selective catalyst for heptane isomerization by
larger surface area than MoO without noble metal after
3
3
reduction at 623 K.19,20 The detailed reduction process to
yield MoO with a large surface area is still under investiga-
tion. The purpose of the present work is to describe the
H
reduction at 623 K.10,11 Both the dehydration and the
2
x
dehydrogenation of propan-2-ol were simultaneously cata-
lyzed by H -reduced MoO .11,12 These results indicate that
relationship between the surface area of H -reduced Pt/MoO
and the extent of reduction. Loading of noble metal onto
2
3
2
3
H -reduced MoO is a porous transition metal oxide with a
2
3
bifunctional catalytic property, and heptane isomerization on
MoO was also reported to improve the catalytic activities for
3
H -reduced MoO proceeds via a conventional bifunctional
heptane isomerization and propan-2-ol dehydration when H
2
3
2
mechanism.
430
reduction was performed at 623 K. The dependency of the
4
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436
This journal is ( The Owner Societies 2001
DOI: 10.1039/b105680h

View Article Online
catalytic activities for the conversions of heptane and propan-
-ol on the reduction degree is also studied in the present
with a 1% O È99% N gas mixture to avoid any strong bulk
oxidation.
2
2
2
work.
Catalytic tests
Experimental
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.
Materials
H
and N2
were puriÐed by passage through a molecular
After H reduction and cooling to reaction temperature in a
2
2
sieve and a Mn/SiO oxygen trap. Heptane and propan-2-ol
H atmosphere, a reactant gas was introduced onto the cata-
2
2
were dried using a molecular sieve prior to use. H MoO with
a purity of 98% was purchased from Kanto Chemical Co. Inc.
Commercially available PtCl É 5H O was used without
further puriÐcation. The MoO used in this study was
lyst bed at partial pressure of 2471 Pa (heptane) or 4825 Pa
2
4
(
propan-2-ol) with H as a complement to atmospheric pres-
2
sure. The composition of effluent gases was analyzed by
means of a FID gas chromatography using a TC-1 glass capil-
lary separation column with temperature-programmed
heating from 303 to 393 K, and TCD gas chromatography
using a unibead 1S separation column at 373 K.
4
2
3
obtained by calcination of H MoO at 673 K for 3 h. The
2
4
0.01 mol.% Pt/MoO was prepared by a conventional impreg-
3
nation method using an aqueous solution of PtCl . 20 g of
4
MoO was added to 20 ml of 7 ] 10~4 M PtCl solution, and
3
4
then the solution was evaporated at 333 K. The obtained
sample was dried overnight at 393 K, followed by calcination
at 673 K for 3 h. The powders were made into pellets, crushed,
and then sieved (30È60 mesh) for charging into a reactor.
Results and discussion
Surface area measurements
The surface areas of H -reduced MoO and Pt/MoO varied
2
3
3
Reduction procedures
markedly with reduction temperature and period. The extent
of reduction depends on reduction conditions. Hence, the
surface areas were plotted as a function of the reduction
degree. Typical results are shown in Fig. 1 where reduction of
A prescribed amount of MoO or 0.01 mol.% Pt/MoO (0.1È
3 3
.5 g) was packed at the central position of a cell, which was
0
made of a Pyrex glass tube with an inner diameter of 6 mm.
The sample was heated to a desired temperature (373È773 K)
at a rate of 5 K min~1 in a stream of H , and was kept at that
temperature for a desired period. It has been reported that
MoO to Mo metal is deÐned to be a reduction degree of
3
1
00%. H -reduced Pt/MoO exhibited almost the same
2
3
2
surface area as the parent Pt/MoO up to a reduction degree
of 30%. The surface area of Pt/MoO was drastically enlarged
by further reduction. The largest surface area was obtained at
3
reduction of MoO and WO was highly a†ected by
3
3
3
reduction conditions, such as mass of sample and Ñow rate of
a reduction degree of 60È70%. The surface area of H -reduced
reducing gas.21h23 Our previous papers showed that the
surface area and the catalytic properties of MoO reduced at
2
MoO was similarly dependent on the extent of reduction as
3
3
that of H -reduced Pt/MoO when MoO was reduced at 573
623 K were inÑuenced by the Ñow rate of H in the reduction
2 3 3
and 623 K. In the case of H -reduced MoO , however, the
2
process.8,9 Hence, the reduction was performed at a H Ñow
2
3
2
rate of 600 ml min~1 g-MoO ~1 in this study. The reduced
3
relationship between the surface area and the reduction degree
was dependent on reduction temperature. The surface area of
samples are denoted as MoO (573-12), Pt/MoO (773-12), etc.
3
3
MoO reduced at 673 K was much smaller than those
The values in the parentheses represent the reduction tem-
perature (K) and period (h).
3
reduced at 573 and 623 K, and was in the range of 40È50 m2
g~1, regardless of the reduction degree. Since Pt/MoO (673-
3
1
2), for which the reduction degree is 63%, has a surface area
Characterization methods
of 247 m2 g~1, the small surface area of MoO reduced at 673
3
K can not be explained by thermal stability.
The surface area was determined from a N adsorption iso-
therm, which was obtained on the sample without exposure to
2
The type of N adsorption isotherm varied with the surface
2
area. Fig. 2 demonstrates the adsorptionÈdesorption iso-
air. The reduced sample was cooled to room temperature
under H2 Ñow. After evacuation for 0.5 h at room tem-
therms of
N
and MoO (623-12), MoO (673-12) and
2
3
3
Pt/MoO (623-12), for which the surface areas were 180 m2
perature, adsorption of N was measured at 77 K using a con-
3
2
g~1, 40 m2 g~1 and 235 m2 g~1, respectively. These reduced
ventional high-vacuum static system. To determine the
pore-size distribution, adsorptionÈdesorption isotherms of N
were measured with an automatic gas-adsorption apparatus
2
(
Sorptomatic 1990, Carlo Erba). Pore-size distribution curves
were obtained by analyzing the adsorption data using the
HorvathÈKawazoe method.
The extent of reduction was calculated from the amounts of
O
consumed in complete reoxidation to MoO . A pulse
2
3
technique was employed in the reoxidation. Since H O was
2
formed in the reoxidation, the concentration of O was moni-
2
tored by TCD gas chromatography using a Porapak N
separation column. The detailed procedures have been
described previously.20
Crystalline phases of H -reduced MoO and Pt/MoO
2
3
3
were determined with X-ray di†raction using Ni-Ðltered
Cu-Ka radiation (Rigaku, Rint-1200). The sample for XRD
measurements was obtained as follows: MoO and Pt/MoO
3
3
were subjected to H reduction at temperatures of 373 to 773
2
Fig. 1 Relation between the surface area and the reduction degree.
MoO reduced at 573 (L), 623 (K) and 673 K (|). Pt/MoO reduced
K for a desired period, followed by Ñowing N for 0.5 h at the
2
3
3
same temperature. After cooling to room temperature under a
at 373 (…), 473 ( Â ), 523 (=), 573 (>), 623 (@), 673 (+), 723 (È) and
773 K ()).
N2 atmosphere, the reduced sample was passivated for 1 h
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436
4431

View Article Online
Fig. 2 AdsorptionÈdesorption isotherms of
N
on H -reduced
2
2
MoO and Pt/MoO . Isotherm: open symbols, adsorption; Ðlled
3
3
symbols, desorption. Sample: (L, …), MoO (623-12); (|, >),
3
MoO (673-12); (K, =), Pt/MoO (623-12).
3
3
samples had a reduction degree of 48%, 57% and 54%,
respectively. The adsorption and desorption isotherms of N
2
on MoO (623-12) and Pt/MoO (623-12) were type I in shape.
3
3
No hysteresis appeared, and the amounts of N adsorbed
2
increased sharply at very low pressures, and changed little at
intermediate pressures. These phenomena are characteristics
of microporous materials. On the other hand, the amounts of
N
adsorbed on MoO (673-12) was raised in proportion to
Fig. 4 Pore-size distributions for H -reduced Pt/MoO . Reduction
2
3
2
3
degree: (A) 39%, (B) 54%, (C) 63% and (D) 93%. (D /D \
the pressure.
v
r
D
volume radius .)
/D
To conÐrm the formation of pores, the pore size distribu-
tion curves were obtained with the HorvathÈKawazoe
parent Pt/MoO . The formation of pores with diameters of 6
method. Fig. 3 shows the pore-size distributions for H -
3
2
to 30 A
Ž
was observed in Pt/MoO (523-12). The value of D /D
reduced MoO and Pt/MoO . MoO (623-12) possessed pores
3
v
r
3
3
3
was raised by an increase in the reduction degree, and reached
a maximum value at a reduction degree of 63%, although no
appreciable di†erence appeared in the pore size distribution.
Further reduction lowered the value of D /D , indicating
rupture of the pores. The pore volumes of Pt/MoO (523-12),
Pt/MoO (623-12), Pt/MoO (673-12) and Pt/MoO (773-12)
were determined to be 0.06, 0.11, 0.13 and 0.07 ml g~1, respec-
tively, from the N adsorption data. These results allow us to
suggest that the formation of pores is a reason for the enlarge-
with diameters of 6 to 30 A
Ž
. Pt/MoO (623-12) was also found
to have pores, although its distribution was somewhat
3
broader than that of MoO (623-12). In contrast, no pores
3
appeared in MoO (673-12). It is obvious from these results
v
r
3
that H -reduced MoO and Pt/MoO with large surface areas
3
3
2
3
3
are porous materials. The pore size distributions for
Pt/MoO (523-12), Pt/MoO (623-12), Pt/MoO (673-12) and
3
3
3
3
3
3
Pt/MoO (773-12), for which the reduction degrees are 39%,
2
5
4%, 63% and 93%, respectively, are displayed in Fig. 4. The
ment of surface area.
surface areas of these reduced samples were 115 m2 g~1, 235
m2 g~1, 247 m2 g~1 and 154 m2 g~1, respectively.
XRD measurements
Pt/MoO (373-12) with a reduction degree of 20% had no
3
pores, and its surface area was comparable to that of the
Changes in the XRD patterns of Pt/MoO on H reduction
3
2
are illustrated in Fig. 5. The di†raction lines due to the hydro-
gen molybdenum bronze phase H MoO were observed in
x
3
Pt/MoO (373-12) with
a
reduction degree of 20%.
3
Pt/MoO (473-12) with a reduction degree of 32% gave no
peak corresponding to the H MoO phase, and was almost
amorphous with respect to XRD. The surface areas of these
3
x
3
Pt/MoO were 9 and 11 m2 g~1, respectively, indicating that
3
the formation of H MoO and the amorphisation were not
x
3
accompanied by an increase in the surface area. The XRD
diagram of Pt/MoO (623-12) with a reduction degree of 54%
3
indicated the appearance of new phases apart from the MoO
2
and Mo metal phase. Two peaks were observed at 2h \ 38.1¡
and 44.3¡. Further reduction promoted the formation of Mo
metal, although the peaks at 2h \ 38.1¡ and 44.3¡ were
observed even at a reduction degree of 93%. In the reduction
of Pt/MoO , the formation of an MoO
2 phase was not
3
detected by XRD.
Fig. 6 shows the XRD patterns of MoO reduced at 573
3
and 673 K. Reduction of MoO at 573 K proceeded via the
3
Fig. 3 Pore-size distributions for MoO (623-12) (L), MoO (673-12)
3
3
(
K) and Pt/MoO (623-12) (|). (D /D \ D
/D
.)
formation of H MoO . MoO (573-24) and MoO (573-48), of
3
v
r
volume radius
x
3
3
3
4432
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436

View Article Online
Fig. 6 XRD patterns of MoO reduced at 573 K (1) and at 673 K
3
(
(
2). Reduction degree in (1): (A) before reduction, (B) 17%, (C) 39%,
D) 46%; in (2): (A) 12%, (B) 21%, (C) 38%, (D) 57%, (E) 95%. (L),
MoO ; (È), H MoO ; (|), MoO ; (@), Mo metal.
3
x
3
2
Fig. 5 XRD patterns of H -reduced Pt/MoO . Reduction degree:
2
3
(
9
A) before reduction, (B) 20%, (C) 32%, (D) 54%, (E) 63% and (F)
3%. (L), MoO ; (È), H MoO ; (@), Mo metal.
3
x
3
ing hydrogen molybdenum bronze, H
MoO , which was
0
.34
3
prepared by introducing MoO into an aqueous solution of
3
which the reduction degrees were 39% and 46%, respectively,
gave di†raction lines at 2h \ 38.1¡ and 44.3¡ along with those
due to the MoO phase. The XRD diagrams of MoO
reduced at 623 K were identical with those at 573 K. As
shown in Fig. 6(B), MoO was reduced to MoO without the
HCl containing chips of Zn metal, with
a
Ñow of
H /hydrocarbon. On the other hand, the treatment of MoO
yielded the mixture of MoO and MoO C phases. The pure
2
3
2
3
2
x y
MoO C was reported to give the di†raction lines at
x y
2h \ 38.2¡ and 43.5¡. As shown in Figs. 5 and 6, H reduction
of Pt/MoO yielded the MoO H phase without the forma-
tion of an MoO phase, while MoO reduced at temperatures
below 623 K consisted of MoO and MoO H phases. The
formation of the MoO H phase is analogous to that of the
MoO C phase. Since the MoO H phase appears only when
3
2
2
formation of H MoO at 673 K. Furthermore, the di†raction
x
3
3
x y
lines at 2h \ 38.1¡ and 44.3¡ were not detected after reduction
at 673 K. These results imply that the reduction process of
MoO di†er with reduction temperature. The H MoO phase
was reported to decompose at 473È573 K in temperature-
programmed reduction.16,19 Thus, the di†erence observed in
Fig. 6 can be understood by taking the thermal stability of the
H MoO phase into consideration. In the presence of Pt, the
2
3
2
x y
3
x
3
x y
x y
x y
the formation of the H MoO phase is detected, the MoO H
x
y
x y
phase seems to be derived from the H MoO phase.
x
y
The XRD results show that H -reduced MoO and
Pt/MoO with large surface areas contained the MoO H
phase. Pt/MoO with a reduction degree of 92% consisted of
the MoO H and Mo metal phases, and its surface area was
x
x
3
3
2 3
H MoO phase was formed during heating to reduction tem-
3
x y
perature so that Pt/MoO was reduced through the formation
3
3
of the H MoO phase, irrespective of reduction temperature.
x
3
x y
This phenomenon is interpreted in terms of hydrogen spill-
over.
about 150 m2 g~1. In contrast, reduction of MoO at 673 K,
3
which yielded a mixture of MoO and Mo metal, enlarged the
2
Delporte and co-workers24 reported that a treatment of
surface area only a little. We propose from these facts that the
MoO with a Ñow of H /hydrocarbon at 623 K yielded a
enlargement of surface area can be induced by the formation
of the MoO H phase, which is derived from the H MoO
3
2
mixture of MoO and molybdenum oxycarbide, MoO C .
2
x y
x y
x
3
This material had a surface area of about 150 m2 g~1, and the
phase.
MoO C phase gave di†raction lines at 2h \ 9.1 and 44.3¡
x y
with Cu Ka. Since MoO treated with pure H at 623 K gave
3
2
Conversions of heptane
Conversion of heptane was carried out at 523 K under atmo-
a di†raction line at 2h \ 44.5¡, Delporte et al. stated that
hydrogen was able to act like a carbon atom to form
MoO H . We deduce on the basis of the reported results that
spheric pressure using H -reduced MoO
and Pt/MoO3
as
x y
2
3
the lines at 2h \ 38.1¡ and 44.3¡ reÑect the formation of
catalysts. The rate of isomerization was determined using data
at the heptane conversion level below 10%. The conversion
level was adjusted by changing the Ñow rate of reactant gas
MoO H phase. Ledoux and co-workers25 have recently
x y
showed that the pure MoO C phase was produced by treat-
x y
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436
4433

View Article Online
Table 1 Product distributions on H -reduced Pt/MoO in the conversion of heptane
2
3
Reduction temp./K
Reduction period/h
573
12
673
12
723
12
773
1
773
6
773
12
Reduction degree (/%)a
48
42.5
63
40.6
79
46.2
67
45.7
76
42.2
93
39.5
%
Conversion
Selectivity (mol/%)
Isomerization
Cracking
96.8
1.0
97.9
1.4
95.2
1.7
96.9
2.6
95.2
4.3
91.6
7.2
Distribution (mol/%)
Isomerization
2
3
-MH
-MH
39.4
44.0
14.8
1.9
39.1
44.5
14.3
2.1
35.2
39.0
20.3
5.5
39.3
40.7
16.5
3.5
37.5
39.5
18.3
4.7
34.1
38.2
21.3
6.4
DMP
Others
Cracking
C
15.5
11.7
28.0
15.8
14.1
1.8
6.6
2.4
4.2
12.4
4.7
29.6
35.5
6.4
1.6
2.7
5.6
3.1
40.5
38.6
5.6
1.4
1.3
4.2
1.9
44.2
38.9
6.9
0.6
1.3
8.6
3.7
39.1
34.0
6.3
0.7
2.7
15.2
8.0
29.3
25.4
5.5
1.4
5.6
1
C
2
C
3
i-C
4
n-C
4
i-C
5
n-C
5
i-C
n-C
2.5
4.7
2.3
1.7
0.6
1.3
1.4
3.5
3.3
7.0
6
6
a The reduction degree was determined from the fresh catalyst.
and/or the weight of catalyst. The H -reduced catalyst
exhibited an almost stable isomerization activity against deac-
tivation. In this study, the catalytic properties were evaluated
using data after a 6 h run.
little when the reduction degree was in the range of 48È79%.
2
Pt/MoO with a reduction degree of 92% exhibited slightly
3
low isomerization selectivity compared with other catalysts.
The distribution of isomerization products was independent of
the reduction degree. Heptane was mainly isomerized to 2-
and 3-methylhexane (MH) in equal amounts. At high conver-
sion levels, dimethylpentanes (DMP) were formed at the
expense of methylhexanes. By contrast, the distribution of
cracking products di†ered from the reduction degree. On
Fig. 7 displays the relationship between the isomerization
activity and the reduction degree. The isomerization activity
was estimated by taking the surface area into consideration,
because the surface area markedly varied with reduction con-
ditions. The isomerization activity of H -reduced Pt/MoO
2
3
was correlated well to the reduction degree. The isomerization
Pt/MoO with a reduction degree of 48%, heptane was
3
activity of Pt/MoO increased with the reduction degree, and
cracked to C ÈC hydrocarbons. The formation of C and C
3
1
6
3
4
the highest isomerization activity appeared at a reduction
hydrocarbons was enhanced at the expense of other hydrocar-
degree of about 70%. The dependency of the isomerization
bons by an increase in the reduction degree. Pt/MoO had the
3
activity of MoO reduced at 573 and 623 K on the extent of
highest selectivity for C and C formation when the
3
3
4
reduction was identical with that of H -reduced Pt/MoO .
reduction degree was 67%. Further reduction lowered the
2
3
On the other hand, reduction of MoO at 673 K improved the
isomerization activity little.
selectivity for C and C
reduction degree of 93% provided similar cracking products
hydrocarbons. Pt/MoO3
with a
3
3
4
The product distributions are demonstrated in Table 1.
Here, the conversion level was adjusted to be about 40% by
varying the Ñow rate of reactant gas and/or the weight of
catalyst. Under the reaction conditions employed, heptane
was selectively isomerized with a minor contribution of crack-
ing and cyclization. The isomerization selectivity changed
to Pt/MoO with a reduction degree of 48%.
Fig. 8 shows the variation in isomerization selectivity with
3
the conversion level of heptane. H -reduced Pt/MoO had the
2
3
same isomerization selectivity as H -reduced MoO when the
2
3
reduction degree of Pt/MoO was in the range of 48È79%.
3
The isomerization selectivity decreased slightly with increasing
conversion level, but the high selectivity was preserved even at
Fig. 7 Relation between the heptane isomerization activity and the
Fig. 8 Dependency of the isomerization selectivity on the conversion
reduction degree. MoO reduced at 573 (L), 623 (K) and 673 K (|).
level of heptane. Pt/MoO with a reduction degree of 48% (È), 63%
3
3
Pt/MoO reduced at 523 (=), 573 (>), 623 (@), 673 (+), 723 (È) and
73 K ()).
(|), 79% (K) and 93% (L). MoO (623-12) with a reduction degree of
3
3
7
48% (…).
4434
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436

View Article Online
high conversion levels on these catalysts. Cracking of the
isomerization products was promoted on Pt/MoO3 with a
reduction degree of 93%, leading to the low isomerization
selectivity at high conversion levels.
the dehydrogenation reaction may proceed mainly on Mo
metal.
Fig. 10 demonstrates the dehydration activity as a function
of reduction degree. Propan-2-ol was dehydrated even on the
parent MoO . Dumesic et al.30 and Belokopytov et al.31 indi-
3
Conversion of propan-2-ol
cated the presence of both Bronsted and Lewis acid sites on
unsupported MoO from IR measurements. This is in agree-
Conversion of propan-2-ol was carried out at 398 K to study
H -reduced Pt/MoO as a bifunctional catalyst. In this study,
3
ment with the reported results. H reduction up to a reduction
2
2
3
degree of about 20% involved a decline in the dehydration
the conversion level of propan-2-ol was adjusted to be below
0% by varying the Ñow rate of reactant gas and/or the
activity. The dehydration activities of H -reduced MoO and
1
2
3
Pt/MoO were improved by further reduction. H -reduced
weight of catalyst. It was reported in previous papers that the
catalytic activity of H -reduced MoO in the conversion of
3
2
Pt/MoO exhibited the highest dehydration activity at a
reduction degree of about 70%. Reduction of MoO at 623 K
was e†ective in enhancing the dehydration activity compared
3
2
3
propan-2-ol was not stable against deactivation.11,14 Hence,
the catalytic activities were compared using data after a 0.5 h
run to minimize the e†ect of catalyst deactivation. Under the
reaction conditions employed, propan-2-ol was converted into
propene and diisopropyl ether (DIPE) by dehydration and to
acetone by dehydrogenation. Since propan-2-ol was much
more selectively dehydrated to propene than to DIPE, the
dehydration activity was evaluated using the rate of propene
formation. The relationships between the reduction degree
and the dehydrogenation and dehydration activities, which
were estimated by taking the surface area into consideration,
were studied.
3
with that at 673 K.
Relation between isomerization activity and dehydration
activity
Alkane isomerization has been proposed to proceed via a
bifunctional mechanism or metallocyclobutane bond-shift
mechanism. Ledoux and co-workers1h3 reported that molyb-
denum oxicarbide, MoO C catalyzed the skeletal isomer-
x y
ization of alkane via
a metallocyclobutane bond-shift
mechanism. In the conversion of heptane, C
carbons are mainly formed as a cracking product on MoO C
and C6
hydro-
1
Fig. 9 shows the dependency of the dehydrogenation activ-
x y
ity on the extent of reduction. In the case of H -reduced
catalyst. As shown in Table 1, H -reduced Pt/MoO with a
2
MoO , the dehydrogenation activity was raised in proportion
3
2
3
reduction degree of about 70% provides C and C hydrocar-
3
4
to the reduction degree, although reduction at 673 K yielded a
bons as the major cracking products, and this catalyst is most
more active catalyst than that at 623 K and 573 K. The dehy-
active for heptane isomerization. The catalytic behavior of H -
2
reduced Pt/MoO in the conversion of heptane is di†erent
3
drogenation activity of Pt/MoO was lowered by H
3
2
reduction up to a reduction degree of 60È80%, and then
increased upon further reduction. Hall and co-workers26,27
have reported that the activity for alkene hydrogenation was
greatly improved by increasing the extent of reduction, and a
sharp upturn in the hydrogenation rate appeared at a
reduction to the average valence of Mo5`. Hercules et al.28,29
suggested that Mo metal and Mo2` were the only active
centers for benzene hydrogenation. The active sites for hydro-
genation are ordinarily capable of catalyzing a dehydroge-
nation reaction. The results of H -reduced MoO can be
from that of MoO C . Since H -reduced MoO and
x y
2
3
Pt/MoO catalyze the dehydration and dehydrogenation of
3
propan-2-ol simultaneously, heptane isomerization is likely to
proceed via a bifunctional mechanism on these catalysts.
As shown in Figs. 7 and 9, the dependency of the isomer-
ization activity on the reduction degree is inconsistent with
that of the dehydrogenation activity. Fig. 11 shows the
relationship between the isomerization activity and the dehy-
dration activity. The isomerization activity was well related to
the dehydration activity, suggesting that the isomerization
activity of H -reduced MoO and Pt/MoO can be controlled
2
3
interpreted in terms of the formation of Mo with lower
valence such as Mo metal, although the dependency of the
dehydrogenation activity on the reduction temperature is still
under investigation. In the case of H -reduced Pt/MoO with
2
3
3
by the ability to act as an acid catalyst. However, there were
two relationships. When compared at the same dehydration
activity, the catalyst with a greater reduction degree exhibited
a higher isomerization activity. The rates of heptane isomer-
2
3
low reduction degrees, Pt will be responsible for the dehydro-
genation. A prolonged period of reduction and reduction at
high temperature, which lead to an increase in the extent of
reduction, seem to promote sintering of Pt, resulting in the
ization on Pt/MoO with reduction degrees of 60% and 79%
3
were 1.1 ] 10~2 and 2.8 ] 10~2 mol h~1 g~1, respectively,
although their dehydration activities were comparable. The
low dehydrogenation activity. On deeply reduced Pt/MoO ,
rates of acetone formation on these catalysts were 3.3 ] 10~3
3
Fig. 9 Relation between the propan-2-ol dehydrogenation activity
Fig. 10 Relation between the propan-2-ol dehydration activity and
and the reduction degree. MoO reduced at 573 (L), 623 (K) and 673
the reduction degree. MoO reduced at 573 (L), 623 (K) and 673 K
3
3
K (|). Pt/MoO reduced at 523 (=), 573 (>), 623 (@), 673 (+), 723
È) and 773 K ()).
(|). Pt/MoO reduced at 523 (=), 573 (>), 623 (@), 673 (+), 723 (È)
3
3
(
and 773 K ()).
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436
4435

View Article Online
tion activities can be enlarged when reduction of MoO pro-
3
ceeds through the formation of a H MoO phase, from which
x
the MoO H phase is derived presumably.
x y
3
Acknowledgements
This work has been supported in part by a Grant-in-Aid for
ScientiÐc Research on Priority Area A of ““New Protium
FunctionÏÏ from the Ministry of Education, Science, Sports
and Culture.
References
1
2
C. P. Huu, M. J. Ledoux and J. Guille, J. Catal., 1993, 143, 249.
E. A. Blekkam, C. P. Huu, M. J. Ledoux and J. Guille, Ind. Eng.
Chem. Res., 1994, 33, 1657.
Fig. 11 Relation between the heptane isomerization activity and the
propan-2-ol dehydration activity. MoO reduced at 573 (L), 623 (K)
3
and 673 K (|). Pt/MoO reduced at 523 (=), 573 (>), 623 (@), 673
3
3
M. J. Ledoux, C. P. Huu, P. Delporte, E. A. Blekkam, A. P. E.
York, E. G. Derouane and A. Fonseca, Stud. Surf. Sci. Catal.,
(
+), 723 (È) and 773 K ()).
1
994, 92, 81.
4
5
6
A. Katrib, P. LeÑaive, L. Hilaire and G. Maire, Catal. L ett., 1996,
and 2.7 ] 10~3 mol h~1 g~1, respectively. Thus, the results
shown in Fig. 11 can not be interpreted in terms of the di†er-
ence in the dehydrogenation activity. Hall and co-workers
reported that cyclopropane isomerization and propene hydro-
genation reactions on H -reduced Mo/Al O were poisoned
38, 95.
A. Katrib, V. Logie, N. Saurel, P. Wehrer, L. Hilaire and G.
Maire, Surf. Sci., 1997, 377–379, 754.
F. Barath, M. Turki, V. Keller and G. Maire, J. Catal., 1999, 185,
1.
2
2 3
with H O.27 The catalytic behavior of H -reduced MoO and
7
8
V. Keller, F. Barath and G. Maire, J. Catal., 2000, 189, 269.
T. Matsuda, Y. Hirata, M. Suzuki, H. Sakagami and N. Taka-
hashi, Chem. L ett., 1999, 873.
2
2
3
Pt/MoO in the isomerization of heptane seems to be some-
what di†erent from that in the conversion of propan-2-ol,
3
9
0
T. Matsuda, Y. Hirata, F. Uchijima, H. Itoh and N. Takahashi,
Bull. Chem. Soc. Jpn., 2000, 73, 1029.
T. Matsuda, H. Shiro, H. Sakagami and N. Takahashi, Catal.
L ett., 1997, 47, 99.
because H O is produced in the conversion of propan-2-ol.
2
The acidity of Mo/Al O has been explained by the induc-
1
2
3
tive e†ect.32,33 Since the average Sanderson electronegativity
of Al O is 3.72, while that of MoO is 3.89, hydroxy groups
on Al O which are near MoO can be made more acidic by
the inductive e†ect. When MoO is reduced to MoO , of
which the electronegativity is 3.53, Mo/Al O becomes less
acidic. Maire and co-workers6,7 recently reported the results
of alkene isomerization on H -reduced MoO /a-Al O . They
proposed from the correlation with XPS measurements that
coordinatively unsaturated sites, such as Mo4` and Mo2`,
will be responsible for dehydrogenation and hydrogenation
reactions, and Bronsted acidity can originate from hydroxy
groups coordinated to Mo5`. The dependencies of the isomer-
ization and dehydration activities on the reduction degree
were very similar to that of the surface area, suggesting an
11 T. Matsuda, Y. Hirata, S. Suga, H. Sakagami and N. Takahashi,
2
2
3
3
3
Appl. Catal. A: General, 2000, 193, 185.
T. Matsuda, Y. Hirata, H. Sakagami and N. Takahashi, Chem.
L ett., 1997, 1261.
3
1
2
3
2
2
3
13 T. Matsuda, Y. Hirata, H. Itoh, H. Sakagami and N. Takahashi,
Microporous Mesoporous Mater., 2001, 42, 337.
14 T. Matsuda, Y. Hirata, H. Sakagami and N. Takahashi, Micro-
porous Mesoporous Mater., 2000, 42, 345.
5
6
7
2
3
2 3
1
1
1
P. A. Sermon and G. C. Bond, J. Chem. Soc., Faraday T rans. 1,
1
976, 72, 730.
G. C. Bond and J. B. Tripathi, J. Chem. Soc., Faraday T rans. 1,
1
976, 72, 933.
P. G. Dickens, R. H. Jarman, R. C. T. Slade and C. J. Wright, J.
Chem. Phys., 1982, 77, 575.
18 R. Erre and J. J. Fripiat, Stud. Surf. Sci. Catal., 1983, 17, 285.
1
2
2
2
2
2
2
2
9
0
1
2
3
4
5
6
T. Matsuda, F. Uchijima, S. Endo and N. Takahashi, Appl. Catal.
A: General, 1999, 176, 91.
important role of the MoO H phase in generating the acid
x y
sites. Although we have no spectroscopic data at present, we
F. Uchijima, T. Takagi, H. Itoh, T. Matsuda and N. Takahashi,
Phys. Chem. Chem. Phys., 2000, 2, 1077.
deduce that interaction of Mon`ÈOH in the MoO H phase
x y
and Mo(n`1)` can create the acid sites that are responsible for
P. Arnoldy, J. C. M. de Jonge and J. A. Moulijin, J. Phys. Chem.,
1985, 89, 4517.
heptane isomerization and propan-2-ol dehydration.
A. Lofberg, A. Frennet, G. Leclercq, L. Leclercq and J. M. Girau-
don, J. Catal., 2000, 189, 170.
Conclusion
T. Ressler, O. Timpe, T. Neisius, J. Find, G. Mestl, M. Dieterle
and R. Schlogl, J. Catal., 2000, 191, 75.
H reduction of Pt/MoO was accompanied by an increase in
2
3
P. Delporte, F. Meunier, C. P. Huu, P. Vennegues, M. J. Ledoux
and J. Guille, Catal. T oday, 1995, 23, 251.
the surface area, which reached the maximum value of 250 m2
g~1 at a reduction degree of 60È70%. The surface area of
MoO reduced at 573 and 623 K was similarly dependent on
the reduction degree as that of H -reduced Pt/MoO . In con-
trast, reduction of MoO at 673 K changed the surface area
C. Bouchy, C. Pham-Huu, B. Heirich, C. Chaumont and M. J.
Ledoux, J. Catal., 2000, 190, 92.
3
E. A. Lombardo, M. Houalla and W. K. Hall, J. Catal., 1978, 51,
256.
2
3
3
little. The formation of pores is likely to be a reason for the
27 E. A. Lombardo, M. Jacono and W. K. Hall, J. Catal., 1980, 64,
1
50.
enlargement of surface area. H -reduced MoO and Pt/MoO
2
3
3
28 R. B. Quincy, M. Houalla, A. Proctor and D. A. Hercules, J.
with large surface areas contained the MoO H phase, which
x y
Phys. Chem., 1990, 94, 1520.
M. Yamada, J. Yasumaru, M. Houalla and D. A. Hercules, J.
Phys. Chem., 1991, 95, 7037.
was derived, presumably, from the H MoO phase. The activ-
2
9
x
3
ity for heptane isomerization was well related to that for
propan-2-ol dehydration, suggesting that the isomerization
activity of H -reduced MoO and Pt/MoO can be controlled
30 T. Kataoka and J. A. Dumesic, J. Catal., 1988, 112, 66.
31 Yu. V. Belokopytov, K. M. Kholyavenko and S. V. Gerei, J.
Catal., 1979, 60, 1.
2
3
3
by the ability to act as an acid catalyst. The dependencies of
the isomerization and dehydration activities on the reduction
degree were very similar to that of the surface area. We
suggest from these results that the isomerization and dehydra-
3
2
W. Suarez, J. A. Dumesic and C. G. Hill, Jr., J. Catal., 1985, 94,
08.
4
3
3
R. T. Sanderson, Chemical Bonds and Bond Energy, Academic
Press, New York, 1976.
4436
Phys. Chem. Chem. Phys., 2001, 3, 4430È4436