Skeletal isomerization of n-pentane over Pt-promoted cesium hydrogen salts of 12-tungstophosphoric acid

Article abstract of DOI:10.1016/S1381-1169(98)00258-1

Two kinds of Pt-promoted Cs_2.5H_0.5PW₁₂O₄₀ (Cs_2.5H_0.5PW₁₂O₄₀ will be denoted by Cs2.5), Pt directly impregnated on Cs2.5 (Pt/Cs2.5) and a physical mixture of Cs2.5 and Pt/Al₂O₃ (Pt + Cs2.5), were used to catalyze the isomerization of n- pentane in the presence of hydrogen at 453-573 K. Cs2.5 showed a high initial activity but deactivated rapidly. Addition of Pt greatly suppressed the deactivation and increased the selectivity to isopentane. High stationary conversion (34.8%) and selectivity (96.9%) were obtained by using Pt + Cs2.5 at a relatively low temperature (453 K) and a low hydrogen pressure (0.05 atm, hydrogen/pentane = 1). Under these reaction conditions, the stationary activity and selectivity of Pt + Cs2.5 were significantly higher than those of Pt-promoted H-ZSM-5 or SO₄/^2-/ZrO₂. It was deduced that the remarkable effect of Pt in suppressing the catalyst deactivation was brought about by activated hydrogen, which were formed on Pt, transferred to Cs2.5, and utilized to remove carbonaceous deposits or their precursors. Increase in the hydrogen pressure decreased the initial activity probably due to a decrease in the concentration of pentenes or pentyl carbenium ion.

Full text of DOI:10.1016/S1381-1169(98)00258-1

Ž
.
Journal of Molecular Catalysis A: Chemical 141 1999 145–153
Skeletal isomerization of n-pentane over Pt-promoted cesium
1
hydrogen salts of 12-tungstophosphoric acid
)
Yanyong Liu, Kyutae Na, Makoto Misono
Department of Applied Chemistry, Graduate School of Engineering, The UniÕersity of Tokyo, Bunkyo-ku, Tokyo 113-8656, Japan
Abstract
Ž
.
Two kinds of Pt-promoted Cs2.5H0.5PW12O40 Cs2.5H0.5PW12O40 will be denoted by Cs2.5 , Pt directly impregnated on
Ž
.
Ž
.
Cs2.5 PtrCs2.5 and a physical mixture of Cs2.5 and PtrAl2O3 PtqCs2.5 , were used to catalyze the isomerization of
n-pentane in the presence of hydrogen at 453–573 K. Cs2.5 showed a high initial activity but deactivated rapidly. Addition
Ž
.
of Pt greatly suppressed the deactivation and increased the selectivity to isopentane. High stationary conversion 34.8% and
Ž
.
Ž
.
selectivity 96.9% were obtained by using PtqCs2.5 at a relatively low temperature 453 K and a low hydrogen pressure
Ž
.
0.05 atm, hydrogenrpentanes1 . Under these reaction conditions, the stationary activity and selectivity of PtqCs2.5
were significantly higher than those of Pt-promoted H–ZSM-5 or SO42yrZrO2. It was deduced that the remarkable effect of
Pt in suppressing the catalyst deactivation was brought about by activated hydrogen, which were formed on Pt, transferred to
Cs2.5, and utilized to remove carbonaceous deposits or their precursors. Increase in the hydrogen pressure decreased the
initial activity probably due to a decrease in the concentration of pentenes or pentyl carbenium ion. q 1999 Elsevier Science
B.V. All rights reserved.
Keywords: Isomerization; n-Pentane; Heteropolyacid; Platinum
1. Introduction
The isomerization of n-pentane is performed
using Friedel–Crafts type catalysts or solid
acids. Friedel–Crafts type catalysts show high
activity even below 373 K, but they have disad-
vantages in separation and corrosivity. Ptr
Al2O3 activated by chlorine-containing com-
pounds and Ptrzeolites are bifunctional cata-
lysts, which are also used commercially. They
are effective for the isomerizations of n-pentane
and n-hexane in the presence of hydrogen usu-
Branched alkanes are becoming important
components of gasoline to enhance the octane-
w
x
number 1,2 . In this respect, the isomerization
of light alkanes catalyzed by solid acids is
w x
beneficial for the production of cleaner fuels 3 .
Carbenium ions formed by strong acid sites are
important intermediates for this reaction 4 . We
studied in this work the isomerization of n-pen-
w x
tane catalyzed by heteropolyacids.
Ž
.
ally at a high reaction temperature )573 K
w
x
5,6 . As lower temperatures are favored by the
equilibrium, the catalysts which consist of noble
)
Corresponding author. Tel.: q81-3-3812-2111; Fax: q81-3-
5802-2949; E-mail: tmisono@hongo.ecc.u-tokyo.ac.jp
metal and various solid acids, such as Ptr
1 Catalysis by Heteropoly Compounds 41. For part 40, see: K.
Inumaru, H. Nakajima, M. Hashimoto, and M. Misono, Nippon
y
SO42 rZrO2 7–9 , Ptrmordenites 10 , Ptrbeta
w
x
w x
w x
w x
Ž
.
zeolites 11 , PtrSiO2 qH–ZSM-5 12 , and Pt
Kagaku Kaishi 1998 390.
1381-1169r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 98 00258-1

(
)
146
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
w
x
modified clay minerals 13 , have been studied.
They catalyze the skeletal isomerization of n-
where the amount of Pt corresponded to 1.0
wt.%. An amount of 1 wt.% PtrAl2O3 was
Ž
Ž
pentane in a medium temperature range 473–
prepared by the impregnation of Al2O3 JRC-
573 K .
Heteropoly compounds are excellent solid
acid catalysts for both gas–solid and liquid–
ALO-4, 167 m2 gy1 with an aqueous solution
.
.
of H2 PtCl6 at 323 K. After the sample was
calcined at 773 K for 3 h in air, it was reduced
at 573 K in a hydrogen flow.
w
x
solid reaction systems 14–17 . H3PW12O40, a
typical heteropolyacid, is a strong acid and ef-
Ž
.
H–ZSM-5 denoted by HZ was obtained
w
x
Ž
Ž
.
fectively catalyzes various reactions 18–20 .
from Na–ZSM-5 HSZ-820NAA Tosoh ,
Ono et al. 21 and Suzuki et al. 22 observed
that Pdx H3y2 x PW12O40rSiO2 catalyzed n-
hexane isomerization in the presence of hydro-
gen. Recently, we found that an acidic Cs salt of
H3PW12O40, namely Cs2.5H0.5PW12O40, was
very active for the skeletal isomerization of
SiO2rAl2O3 s 23.2, surface area s 322 m2
w
x
w x
gy by an ion-exchange method. Na–ZSM-5
1
.
was treated with an aqueous solution of
NH4NO3 to form NH4–ZSM-5, followed by
drying at 383 K. NH4–ZSM-5 was added to an
aqueous solution of H2 PtCl6. After the solution
was stirred at 353 K for 3 h, the solid was
filtrated and washed with deionized water, and
w
x
n-butane 23,24 and the stationary activity was
enhanced greatly by the addition of platinum at
w
x
low hydrogen pressures 25,26 .
then dried at 383 K and calcined at 773 K for 3
y
In the present study, Pt-promoted Cs2.5 cata-
lysts have been applied to the isomerization of
n-pentane. By comparing the results with those
h. SO42 rZrO2 denoted by SZ was prepared
Ž
.
w
x
Ž
.
by a method in the literature 28 . Zr OH ,
which was obtained by the hydrolysis of ZrOCl2
4
of Pt-promoted SO42 rZrO2 and Pt-promoted
solution by NH3, was treated with an aqueous
y
Ž
.
H–ZSM-5, the catalytic features of Pt-promoted
Cs2.5 for n-pentane isomerization have been
discussed.
solution of H2SO4 1 N , and the resulting solid
was calcined at 773 K for 5 h in air.
y
PtrSO42 rZrO2 abbreviated as PtrSZ was
Ž
.
prepared as in the literature by the impregnation
2
y
Ž
.
of SO4 rZr OH with an H2 PtCl6 aqueous
4
2. Experimental
solution followed by drying at 383 K and calci-
w x
nation at 773 K in air 8 .
2.1. Catalysts
The physically mixed catalysts of PtrAl2O3
and either of Cs2.5, HZ or SZ were prepared as
follows. First, 1 wt.% PtrAl2O3 was mixed
with the same amount of Cs2.5, HZ or SZ, and
ground in a mortar. Then the mixture was
pressed to a pellet, ground to granules, and
sieved to 24–60 mesh. These catalysts are de-
noted by PtqCs2.5, PtqHZ and PtqSZ, re-
spectively.
Ž
.
Cs2.5H0.5PW12O40 abbreviated as Cs2.5 was
prepared by the titration of the aqueous solution
of H3PW12O40 by an aqueous solution of
Cs2CO3 as in the previous work 27 . PtrCs2.5
was prepared by the following step. An aqueous
w
x
y
3
Ž
.
solution of H2 PtCl6 0.04 mol dm
dropwise to an aqueous solution of H3PW12O40
was added
y
3
Ž
.
0.06 mol dm
at that temperature. Then, an aqueous solution
at 323 K, and held for 15 min
2.2. Skeletal isomerization of n-pentane
y
3
Ž
.
of Cs2CO3 0.08 mol dm
was added to the
PtrH3PW12O40 solution at 323 K with a rate of
about 0.6 dm3 miny1, and held overnight at
room temperature. The colloidal solution ob-
tained was evaporated to dryness at 323 K. The
molar ratio of Pt:Cs:PW12O40 was 0.18:2.5:1.0,
The isomerization of n-pentane was per-
Ž
formed in a flow reactor Pyrex, 12 mm in outer
3
y1
.
diameter . The total flow rate was 10 cm min
and 1 g of the catalysts were used. Prior to the
reaction, PtrCs2.5 was treated at 573 K in a

(
)
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
147
Ž
.
Fig. 1. Time courses of n-pentane isomerization at 453 K.
I
Ž
.
Fig. 2. Conversions of n-pentane with time at 453 K.
v
Ž .
^
Ž .
`
Ž .
Cs2.5,
gas: n-pentane, 0.05 atm; H2 , 0.05 atm; N2 balance. WrF, 40 g h
moly1
PtrAl2O3,
PtrCs2.5, and
v PtqCs2.5. Feed
Ž
.
Ž .
PtqCs2.5, B PtqHZ, ' PtqSZ. Feed gas: n-pentane, 0.05
atm; H2 , 0.05 atm; N2 balance. WrF, 40 g h moly1
.
.
hydrogen flow for 2 h, and other catalysts were
pretreated at the reaction temperature for 2 h in
a hydrogen flow. The reaction products were
initial stage of the reaction. The selectivity in-
creased with the reaction time from the initial
selectivity of 20.2 to 100% after 4 h. The
catalytic activity of PtrAl2O3 was very low
under these reaction conditions. When Pt was
added to Cs2.5, high conversions were main-
Ž
.
analyzed with a FID GC Hitachi GC-163
equipped with an Al2O3rKCl fused silica col-
umn.
Ž
.
tained PtrCs2.5: 27.1%, PtqCs2.5: 34.8% ,
and high selectivities were observed at the sta-
tionary state. PtqCs2.5 showed the highest
stationary conversion among the four catalysts.
PtqCs2.5 is compared with PtqHZ and
PtqSZ in Fig. 2. The initial activity of PtqSZ
was higher than PtqCs2.5, but it deactivated
rapidly. PtqCs2.5 gave the highest stationary
conversion. In Table 2, the catalytic activities
3. Results
Fig. 1 shows typical time courses of the
conversion of n-pentane over Cs2.5, PtrCs2.5,
PtrAl2O3, and PtqCs2.5 at 453 K with 0.05
atm of hydrogen. The selectivities are summa-
rized in Table 1. The initial activity of Cs2.5
Ž
.
was very high more than 40% conversion , but
the conversion decreased considerably at the
Table 1
Conversions and selectivities of isomerization of n-pentane at 453 K
Ž
.
Catalyst
Time
on stream
Conversion
Selectivity %
Ž
.
%
C3
C4
iyC5
C6q
Cs2.5
5 min
4 h
42.1
1.1
9.4
0
67.4
0
20.2
100
2.4
0
PtrAl2O3
PtrCs2.5
5 min
4 h
5 min
4 h
5 min
4 h
2.1
1.4
43.2
27.1
29.9
34.8
6.2
3.3
0.5
0.3
0
37.4
25.9
8.3
2.1
0
45.7
62.3
87.0
96.4
100
7.0
4.9
4.2
1.1
0
PtrAl2O3 qCs2.5
0
1.7
96.9
1.4
Feed gas: n-pentane, 0.05 atm; H2 , 0.05 atm; and N2 balance.
WrF, 40 g h moly1 ; catalyst, 1 g.

(
)
148
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
Table 2
Conversions and selectivities of skeletal isomerization of n-pen-
tane at 453 K
Ž
.
Catalyst
Conversion Selectivity %
Ž .
%
C3 C4 iyC5 C6q
PtrAl2O3 qCs2.5
34.8
19.2
0.0 1.7 96.9
0.2 3.9 93.3
3.7 9.1 79.4
2.3 6.8 81.8
0.0 1.8 97.6
0.0 0.4 99.6
1.4
2.6
7.8
3.1
0.8
0.0
PtrAl2O3 qHZSM-5
PtrAl2O3 qSO42yrZrO2 9.4
PtrSO42yrZrO2
5.2
64.7
6.3
PtrAl2O3 qHZSM-5a
PtrSO42yrZrO2b
Feed gas: n-pentane, 0.05 atm; H2 , 0.05 atm; and N2 balance;
WrF, 40 g h moly1
aReaction temperature, 523 K. n-Pentane, 0.1 atm; H2 , 0.9 atm;
.
WrF, 10 g h moly1 ; Pt 2.5% rAl2O3, 0.2 g; and H–ZSM-5, 0.8
Ž
.
w
x
g. Compare with the date in Ref. 29 .
Fig. 3. Effect of WrF on conversion for n-pentane isomerization
over PtqCs2.5 at 453 K. Feed gas: n-pentane, 0.05 atm; H2 , 0.05
atm; N2 balance.
bReaction temperature: 523 K. n-Pentane, 0.3 atm; H2 , 0.7 atm;
WHSV, 17.3 hy1 ; and Pt, 0.5 wt.%. Compare with the data in
w x
Ref. 8 .
Ž
.
x is defined by Csx H3yx PW12O40 . The sta-
and the product distributions of Pt-promoted
solid acid catalysts are summarized. The selec-
tivity to isopentane of PtqCs2.5 was compara-
ble to that of PtqHZ and higher than that of
PtqSZ. PtrSZ was less active than PtqSZ, as
in the case of n-butane isomerization 30 . Thus
PtqCs2.5 is an excellent catalyst for skeletal
isomerization of n-pentane even under a low
pressure of hydrogen and at a relatively low
tionary activity showed a maximum at xs2.5,
and the selectivity was almost constant except
for a significant decrease for xs3.
Results for a series of PtqCs2.5 catalysts
having different mixing ratios, where the amount
of Cs2.5 was kept constant and the partial pres-
sure of n-pentane and hydrogen were 0.05 atm,
are plotted in Fig. 6. With an increase in the
content of PtrAl2O3, the deactivation was re-
markably suppressed. The effects are further
demonstrated in Fig. 7. When a small amount of
w
x
Ž
temperature. Data in Table 2 last two lines in
.
the table also demonstrate that PtqHZ and
PtrSZ used in the present study were very
similar to those reported in the literature 8,12 ,
since they showed almost the same activities
and selectivities as those in the literature under
the same reaction conditions.
w
x
In Fig. 3, it was confirmed that the stationary
conversion for PtqCs2.5 increased linearly
with WrF, the ratio of the catalyst weight to the
flow rate of pentane. The effects of the pretreat-
ment temperature on stationary conversion and
selectivity are shown in Fig. 4. In the range of
453–573 K, the conversion and selectivity were
nearly constant. The conversion decreased
greatly with the pretreatment temperature above
573 K and the selectivity increased from 97 to
100%. Fig. 5 shows the effects of the Cs content
of the heteropolyacid component in PtqCs x
Fig. 4. Effects of pretreatment temperature on conversion and
selectivity for n-pentane isomerization at 453 K over PtqCs2.5.
Ž
v
.
Ž .
Conversion, ` selectivity. Feed gas: n-pentane, 0.05 atm;
H2 , 0.05 atm; N2 balance. WrF, 40 g h moly1
.

(
)
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
149
Ž
.
Fig. 5. Effect of Cs content of n-pentane isomerization over 1%
Ž
.
Ž .
PtrAl2O3 qCsx H3yx PW12O40. v Conversion, ` selectivity.
Feed gas: n-pentane, 0.05 atm; H2 , 0.05 atm; N2 balance. WrF,
40 g h moly1 ; reaction temperature, 453 K.
Fig. 7. Effect of Pt loading on n-pentane isomerization over
PtqCs2.5 at 453 K. Feed gas: n-pentane, 0.05 atm; H2 , 0.05 atm;
N2 balance. Cs2.5: 0.5 g, total flow rate: 10 cm3 miny1
Ž .
. v
Ž
. Ž
.
Ž
Stationary conversion after 4 h , ` stationary selectivity after
.
4 h .
PtrAl2O3 was added to Cs2.5, the stationary
conversion and selectivity increased greatly, but
for the amount of PtrAl2O3 greater than 0.2 g,
the conversion and selectivity remained con-
Ž
n-pentane pressure i.e., first order in n-pen-
.
tane at 453 K at 0.05 atm of hydrogen pressure.
However, the rate at the stationary state solid
squares tended to level off at higher pressures
of n-pentane )0.2 atm .
Fig. 9 shows the time courses of conversion
of the n-pentane catalyzed by PtqCs2.5 at 453
K at various hydrogen pressures. The initial
activity was very high in nitrogen, but the deac-
Ž
stant. As for the initial stage not shown in Fig.
.
Ž
7 , the conversion remained almost constant but
.
the selectivity increased greatly with increasing
of PtrAl2O3.
Fig. 8 shows the dependences of the reaction
rates on the partial pressure of n-pentane. The
Ž
.
Ž
.
initial rate solid circles was proportional to the
Ž
.
Fig. 6. Time courses of n-pentane isomerization over 1%
Ž
.
Ž .
PtrAl2O3 qCs2.5. Cs2.5, 0.5 g; PtrAl2O3: I 0 g, B 0.05 g,
Fig. 8. Pressure dependence of n-pentane isomerization over
Ž
.
Ž .
Ž
.
Ž
. Ž
.
`
0.1 g, v 0.5 g. Feed gas: n-pentane, 0.05 atm; H2 , 0.05
PtqCs2.5 at 453 K. v Initial rate 5 min ; B stationary rate
atm; N2 balance. Total flow rate, 10 cm3 miny1 ; reaction temper-
ature, 453 K.
4 h ; partial pressure of H2 , 0.05 atm; N2 balance; catalyst, 1 g;
Ž
.
WrF, 40 g h moly1
.

(
)
150
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
with the hydrogen partial pressure at each tem-
perature, but the conversions at the stationary
state showed maxima at certain hydrogen partial
pressures. The maxima were observed at higher
hydrogen pressures as the reaction temperature
Ž
.
increased Fig. 10D . The highest stationary
yield was obtained at 493 K in 0.2 atm of
hydrogen pressure; the conversion was 62% and
Ž
the selectivity 91%. In the case of Cs2.5 without
.
Ž
.
Pt , hydrogen pressure 0–0.5 atm little influ-
enced the reaction. Under high hydrogen pres-
Ž
.
sure 0.5 atm at 493 K, stationary conversion
Ž
.
was higher for PtqSZ 63% than PtqCs2.5
Fig. 9. Time courses of the isomerization of n-pentane over
Ž
Ž
.
52% , but the selectivity was lower for PtqSZ
Ž
.
Ž
.
Ž
.
PtqCs2.5 at 453 K. Feed gas: H2: ` 0 atm, v 0.02 atm, I
.
Ž
.
92% than PtqCs2.5 98% .
Ž .
B
Ž .
0.05 atm,
0.20 atm,
^ 0.50 atm; n-pentane, 0.05 atm; N2
balance. WrF, 40 g h moly1
.
Fig. 11 shows the temperature dependence of
n-pentane isomerization over PtqCs2.5. Com-
tivation was serious. When the hydrogen pres-
sure increased, the deactivation was suppressed
gradually, while the initial conversion de-
creased. Pressure dependences of the reaction
rate on hydrogen for PtqCs2.5 at 423, 453 and
493 K are given in Fig. 10. It was found that
Ž
parison of the conversions at 5 min open cir-
.
Ž
.
cles and 4 h solid circles show that the deacti-
vation took place above 493 K with 0.05 atm
hydrogen pressure, while there was little deacti-
vation even at 573 K when 0.5 atm hydrogen
was present. By-products were C3, C4, C6 and
C7 below 493 K. C1 and C2 were observed at
493 K and increased significantly above 533 K.
Ž
.
both the initial after 5 min and stationary
Ž
.
after 4 h selectivities increased with the hydro-
gen pressure. The initial conversions decreased
Fig. 11. Effect of reaction temperature on n-pentane isomerization
Ž
.
Ž .
Fig. 10. Effect of H2 partial pressure on n-pentane isomerization
over PtqCs2.5 in the H2 pressure of: A 0.05 atm, B 0.5 atm.
over PtqCs2.5.
'
423 K,
. Ž .
v
453 K,
B
493 K.
Ž
A
initial
. Ž .
n-Pentane, 0.05 atm; WrF, 40 g h moly1. ` Initial conversion
Ž
.
Ž
.
Ž
.
Ž
.
Ž .
Ž
Ž
Ž
Ž
. Ž
.
Ž
. Ž .
selectivity after 5 min ; B stationary selectivity after 4 h ; C
5 min , v stationary conversion 4 h , I selectivity to iyC5
Ž
. Ž
.
Ž
.
. Ž
.
B
Ž
. Ž
. Ž
. Ž
.
initial conversion after 5 min ; D stationary conversion after 4
.
5 min ,
selectivity to C3 qC4 5 min ,
'
selectivity to
.
h ; n-pentane, 0.05 atm; N2 balance; WrF, 40 g h moly1
C1qC2 5 min , l selectivity to C6 qC7 5 min .
. Ž
. Ž
.
Ž

(
)
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
151
4. Discussion
drogenolysis was not negligible in the case of
w
x
n-butane isomerization 25,26 . Therefore, it
may be stated that the remarkable effects of Pt
are to increase the selectivity to isopentane and
suppress the deactivation. Pt catalyzes the hy-
drogenation of coke or coke precursor covering
4.1. Roles of surface acidity
H3PW12O40 is so far the strongest acid among
heteropoly compounds 31 . Cs2.5 is usually the
most active catalyst for the surface-type cataly-
w
x
w
x
acid sites by supplying hydrogen 32 . This may
explain the suppression of deactivation.
w
x
sis among the Cs salts of H3PW12O40 16,17 .
This is because the acid strength of the Cs2.5 is
approximately the same as H3PW12O40 and its
surface area is very high. In consequence, the
amount of surface proton of Cs2.5 for unit
weight is much higher than that of H3PW12O40.
The activity change of PtqCs x shown in Fig.
5 correlates well with the amount of surface
As shown in Fig. 1, although the initial activ-
ity of PtrCs2.5 was higher than that of Ptq
Cs2.5 because the content of Cs2.5 was higher
Ž
.
ca. 2:1 . The deactivation was greater for
PtrCs2.5 possibly due to the dispersion of Pt
was lower for PtrCs2.5. The amount of hydro-
gen adsorbed per Pt for PtrAl2O3 was more
than twice that for PtrCs2.5. Little deactivation
results in the high activity of PtqCs2.5 after 4
h.
w
x
acid sites reported previously 16,17 , indicating
that the activity is primarily determined by the
surface acidity.
The structure of Cs2.5 is stable and the
amount of the water of crystallization is approx-
imately null in the range of 373–573 K. But the
lattice oxygens are removed with protons in the
form of water above 613 K resulting in the loss
As shown in Figs. 9 and 10, the partial
pressure of hydrogen showed a great influence
on the reaction in the presence of both PtrAl2O3
and Cs2.5. As discussed above, hydrogen sup-
pressed the deactivation probably by hydro-
genating the coke precursor or coke which
causes the deactivation of catalysts. The nega-
tive dependence of the initial rate and deactiva-
w
x
of acidity 24 . This explains the results shown
in Fig. 4; the activity of PtqCs2.5 was con-
stant when the pretreatment temperature was
below 573 K, and decreased greatly when the
temperature was above 613 K.
Ž
.
tion on the hydrogen pressure Fig. 9 may be
explained by the decrease in the concentration
Ž
.
of pentenes in bifunctional catalysis see below .
The hydrogen pressure for the highest stationary
conversion increased with the reaction tempera-
4.2. Roles of Pt and H2
Ž
.
The addition of PtrAl2O3, which was almost
inactive, to Cs2.5 did not enhance the initial
activity but the stability was significantly im-
ture Fig. 10D . This is explained in a similar
way. Since the pentene concentration would
increase at high temperature, a higher hydrogen
pressure is required to suppress the coke forma-
tion.
Ž
.
proved Fig. 1 . When the amount of PtrAl2O3
mixed was low, the deactivation was not sup-
pressed completely, but with a sufficient amount
of PtrAl2O3 the activity rather increased in the
initial stage and a high stationary activity was
4.3. Reaction mechanism
Ž
.
maintained Fig. 6 . Fig. 7 shows that 0.2 g of
PtrAl2O3 was sufficient to eliminate the deacti-
vation. The selectivity was high and nearly con-
stant even when the amount of PtrAl2O3 was
high, which shows the hydrogenolysis on
PtrAl2O3 is not significant in this case under
these reaction conditions. Note that the hy-
As high performance of pentane isomeriza-
tion was obtained by using Pt-promoted Cs2.5
Ž
.
PtqCs2.5 and PtrCs2.5 , the mechanism of
the reaction is an interesting subject. The main
aim of the present study is to make clear the
catalytic feature of the heteropoly catalysts and
there is not sufficient experimental evidence to

(
)
152
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
conclude the mechanism unambiguously. How-
ever, the following discussion may be made to
understand the present results. It is certain as
discussed above that the acidity of Cs2.5 plays
an important role and pentyl carbenium ions
formed on acid sites are the key intermediates.
There are several possible mechanisms for the
formation of 2-pentyl cation and the conversion
of isomerized 2-methylbutyl cation to isopen-
tane. As for the formation of carbenium ions,
Rather small influence of the presence of Pt on
Ž
.
the initial activity Fig. 6 may not favor the
classical bifunctional mechanism, in which an
increase in the activity is expected, and hence
favors the spill-over mechanism. In either case,
the presence of Pt and hydrogen must increase
the stationary activity and selectivity, by the
suppression of olefin concentration by hydro-
genation. This prevents the polymerization or
coke formation. Another important effect is that
Ž .
Ž
.
there are three possible routes: i protonation to
the spilled-over active hydrogen in any form
form carbonium ion followed by its transforma-
hydrogenates the coke or coke precursors on the
Ž .
tion to carbenium ion and hydrogen; ii hydride
solid acids and prevent deactivation.
abstraction of pentane by Lewis acid site; and
Ž
.
iii dehydrogenation of pentane to pentene and
4.4. Side reactions
subsequent protonation. After the isomerization
of 2-pentyl cation to 2-methylbutyl cation, hy-
dride transfer to it from n-pentane or hydrogen
produces isopentane, or deprotonation of the
cation to isopentene followed by its hydrogena-
tion on Pt forms isopentane. Bimolecular mech-
anism is another possible path, and this also
proceeds via carbenium ions.
There are two possible roles of Pt. One is the
formation of pentene by dehydrogenation of
pentane. Once pentene is formed, its isomeriza-
tion to isopentene can proceed via pentyl cations
on acid sites and the isomerized 2-methylbutene
is easily hydrogenated to isopentane. This is the
well-known bifunctional mechanism 33,34 .
Negative effect of hydrogen on the initial activ-
ity favors this mechanism Fig. 9 . Another
possibility is the spill-over of hydrogen atoms
from Pt to the surface of solid acid and forms
protons there. Hattori assumes that protons and
The main side reactions of isomerization of
n-pentane are usually disproportionation and hy-
drogenolysis. Disproportionation reaction is
alkylation-cracking via a C10 intermediate and
is a part of ‘bimolecular mechanism’ 36,37 ,
where the main by-products are C3 qC7 or
C4 qC6. C6 and C7 are more reactive and may
be cracked to smaller fragments. The hy-
drogenolysis converts C5 to C1 qC4 or C2 q
C3. This type of side reactions is likely depen-
dent on the catalysts and reaction conditions.
w
x
Ž
.
The formation of C6 over Cs2.5 alone Table 1
w
x
suggests the presence of disproportionation re-
action by the bimolecular mechanism. The high
selectivity to isopentane of PtrCs2.5 and Ptq
Cs2.5 in the presence of sufficient hydrogen
Ž
.
Ž
.
Table 1 indicates that the disproportionation
reaction or the bimolecular mechanism is not
electrons formed are trapped on solid acid in the
important under these reaction conditions.
y
case of PtrSO42 rZrO2 Eq. 1 35 . Iglesia et
The contribution of the two side reactions
may be evaluated by product distributions. For
PtqCs2.5, C6 and C7 were observed at 0.05
atm of hydrogen above 493 K, but not observed
Ž
Ž .. w x
w x
w x
al. 7 and Zhang et al. 29 suggested the
Ž .
Ž .
following mechanisms, Eqs. 2 and 3 , respec-
tively.
Ž
.
at 0.5 atm of hydrogen Fig. 11 . This indicates
that the disproportionation reaction is sup-
pressed by hydrogen. Hydrogenolysis reaction
occurs in the high temperature region, as C1 and
C2 were observed above 533 K both with 0.05
and 0.5 atm of hydrogen.
PtyH™PtqHqqe PtrSO2yrZrO2
1
Ž .
Ž
.
4
2Pt–H™Pty–HqHq PtrSO2yrZrO2
2
Ž .
Ž
.
4
2Pt–H™PtqHqqHy PtrHZSMy5
3
Ž .
Ž
.

(
)
Y. Liu et al.rJournal of Molecular Catalysis A: Chemical 141 1999 145–153
153
w x
Ž .
D. Seddon, Catal. Today 15 1992 1.
H. Hattori, O. Takahasi, M. Takagi, K. Tanabe, J. Catal. 68
3
w x
4
4.5. Comparison of Õarious solid acid catalysts
Ž
.
1981 132.
It may be accepted that the skeletal isomer-
ization of n-pentane over solid acids and Pt-pro-
moted solid acids proceeds via carbenium ions
present on acid sites and the acidic property is
an important factor in controlling the reaction.
When a noble metal is added to a solid acid,
w x
Ž .
5
w x
6
J.H. Sinfelt, Adv. Chem. Eng. 5 1964 37.
Ž .
F. Ribeiro, C. Marcilly, M. Guisnet, J. Catal. 78 1982 275.
E. Iglesia, S.L. Soled, G.M. Kramer, J. Catal. 144 1993
w x
7
Ž
.
238.
w x
Ž
.
8
K. Ebitani, J. Konishi, H. Hattori, J. Catal. 130 1991 257.
J. Salmones, R. Licona, J. Navarrete, P. Salas, Catal. Lett. 36
w x
9
Ž
.
1996 135.
w
w
w
x
Ž
.
10 K. Chao, H. Wu, L. Leu, Appl. Catal. A: General 143 1997
Ž
the balance between acidity strength and
223.
.
x
11 H. Liu, G.D. Lei, W.M.H. Sachtler, Appl. Catal. A: General
amount , metal amount and hydrogen partial
Ž
.
137 1996 167.
pressure must be important for the optimum
performance of the isomerization. Strong acid
site would exhibit a high activity, but it tends to
be deactivated easily by carbonaceous deposits.
The titration by Hammett’s indicators showed
that acid sites stronger than y16 in H0 function
x
12 K. Fujimoto, K. Maeda, K. Aimoto, Appl. Catal. A: General
Ž
.
91 1992 81.
w
w
w
w
w
x
Ž
.
13 V.N. Parulekar, J.W. Hightower, Appl. Catal. 35 1987 249.
x
Ž
Ž
Ž
.
.
14 M. Misono, Catal. Rev. Sci. -Eng. 29 1987 269.
x
15 M. Misono, Catal. Rev. Sci. -Eng. 30 1988 339.
. Ž .
x
16 M. Misono, T. Okuhara, Chemtech 23 11 1993 23.
x
17 T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. Rel. Subj.
Ž .
41 1996 113.
w
x
are present on the surface of SZ 28 . NH3 TPD
w
w
w
w
x
Ž
.
.
18 I.V. Kozhevnikov, Russ. Chem. Rev. 56 1987 811.
w
x
also indicated its strong acidity 31,38 . Accord-
ingly, SZ is very active, but deactivates fast
x
Ž
19 I.V. Kozhevnikov, Russ. Chem. Rev. 62 1993 473.
x
Ž
.
20 H. Hayashi, J.B. Moffat, J. Catal. 83 1983 192.
w
x
x
39 . Hence, a high hydrogen pressure is needed
to suppress the deactivation.
21 Y. Ono, M. Taguchi, S. Suzuki, T. Baba, in: B. Imelik, C.
Ž
.
Naccache, G. Courdurier, Y.B. Taarit, J.C. Eds. , Vedrine
Catalysis by Acids and Bases, Elsevier, Amsterdam, 1985, p.
167.
According to the temperature-programmed
desorption of NH3, the acid strength of
H3PW12O40 was lower than that of SZ, but
w
w
w
x
Ž
.
22 S. Suzuki, K. Kogai, Y. Ono, Chem. Lett. 1984 699.
x
Ž
.
23 K. Na, T. Okuhara, M. Misono, Chem. Lett. 1993 1141.
x
24 K. Na, T. Okuhara, M. Misono, J. Chem. Soc. Faraday
w
x
higher than that of HZ 31,38 . Another interest-
ing feature of Cs2.5 is that the acid strength is
Ž
.
Trans. 91 1995 367.
w
w
x
25 K. Na, T. Okuhara, M. Misono, J. Chem. Soc. Chem.
Ž .
Commun. 1993 1422.
w
x
nearly uniform 31,38 . It is therefore probable
that the coke formation is not so fast as in the
case of SZ and the deactivation due to coke is
more easily suppressed than the case of PtqSZ
or PtrSZ. The deactivation was smaller also for
PtqHZ, but the activity was low. The lower
activity of PtqHZ is explained by the weaker
acid strength of HZ than Cs2.5. Thus the uni-
formity and moderate strength of the acid sites
of Cs2.5 are most probably the reasons why
Pt-promoted Cs2.5 showed high activity and
good stability.
x
26 T. Okuhara, K. Na, M. Misono, in: Y. Izumi, H. Arai, M.
Ž
.
Iwamoto Eds. , Science and Technology in Catalysis 1994,
Kodansha, 1995, p. 245.
w
w
w
w
w
w
x
27 S. Tatematsu, T. Hibi, T. Okuhara, M. Misono, Chem. Lett.
Ž
.
1984 865.
x
Ž
.
28 M. Hino, K. Arata, J. Chem. Soc. Chem. Commun. 1980
851.
x
29 A. Zhang, I. Nakamura, K. Aimoto, K. Fujimoto, Ind. Eng.
Ž
.
Chem. Res. 34 1995 1074.
x
30 J.C. Yori, M.A. D’Amato, G. Costa, J.M. Parera, J. Catal.
Ž
.
153 1995 218.
x
31 T. Okuhara, T. Nishimura, H. Watanabe, M. Misono, J. Mol.
Ž
.
Catal. 74 1992 247.
x
32 F. Garin, D. Andriamasinoro, A. Abdulsamad, J. Sommer, J.
Ž
.
Catal. 131 1991 199.
w
w
w
w
w
w
x
Ž
.
33 P.B. Weisz, E.W. Swegler, Science 126 1957 31.
x
Ž
.
34 F. Ribeiro, C. Marcilly, M. Guisnet, J. Catal. 78 1982 267.
x
Ž
.
35 H. Hattori, Stud. Surf. Sci. Catal. 77 1993 69.
References
x
Ž
.
36 N.Y. Chen, Stud. Surf. Sci. Catal. 28 1986 653.
x
Ž
.
37 G.A. Fuentes, B.C. Gates, J. Catal. 76 1982 44.
w x
1
J.G. Speight, The Chemistry and Technology of Petroleum,
Marcel Dekker, New York, 1980, p. 268.
x
38 T. Nishimura, T. Okuhara, M. Misono, Appl. Catal. 73
Ž
.
1991 L7.
w x
2
Ž .
M. Chiron, Stud. Surf. Sci. Catal. 30 1987 1.
w
x
Ž
.
39 J.C. Yori, J.C. Luy, J.M. Parera, Appl. Catal. 46 1989 103.