Mechanistic study on skeletal isomerization of n-butane using 1,4- 13C2-n-butane on typical solid acids and their Pt-promoted bifunctional catalysts

Article abstract of DOI:10.1016/j.molcata.2003.08.021

The reaction mechanism of skeletal isomerization of n-butane over typical solid acids such as Cs_2.5H_0.5PW₁₂O ₄₀, sulfated ZrO₂, and WO₃/ZrO₂ and their Pt-promoted catalysts have been studied by the use of 1,4- ¹³C₂-n-butane. Isotopic distributions in product isobutane and reactant n-butane were quantitatively analyzed by field-ionization mass spectrometry (FI-MS). On these solid acids, below 423K, isobutane consisted of ¹³C₀-¹³C₄ isotopes, whose distributions were close to the corresponding binomial distributions, with the reactant n-butane being mainly ¹³C ₂-n-butane. This indicates that the isomerization proceeded with an intermolecular rearrangement through a bimolecular mechanism. In the temperature range of 493-523K over these solid acids, the isotopic distributions in isobutane deviated from the binomial distributions; the fractions of ¹³C₂-isobutane were greater than those expected from the binomial distributions, showing that an intramolecular (monomolecular) rearrangement became significant. Further, the product isobutane was found to consist of exclusively ¹³C₂- isobutane over these Pt-promoted catalysts in the presence of H₂, demonstrating the operation of a monomolecular mechanism on the bifunctional catalysts.

Full text of DOI:10.1016/j.molcata.2003.08.021

Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
Mechanistic study on skeletal isomerization of n-butane
using 1,4-13C2-n-butane on typical solid acids and
their Pt-promoted bifunctional catalysts
Tsuneo Echizena, Tetsuo Suzukia, Yuichi Kamiyab, Toshio Okuharaa,∗
a
Graduate School of Environmental Earth Science, Hokkaido University, Sapporo 060-0810, Japan
b
Japan Science and Technology Corporation, 4-1-8 Honcho, Kawagachi 332-0012, Japan
Received 17 June 2003; received in revised form 28 July 2003; accepted 18 August 2003
Abstract
The reaction mechanism of skeletal isomerization of n-butane over typical solid acids such as Cs2.5H0.5PW12O40, sulfated ZrO2, and
WO3/ZrO2 and their Pt-promoted catalysts have been studied by the use of 1,4-13C2-n-butane. Isotopic distributions in product isobutane and
reactant n-butane were quantitatively analyzed by field-ionization mass spectrometry (FI-MS). On these solid acids, below 423 K, isobutane
consisted of 13C0–13C4 isotopes, whose distributions were close to the corresponding binomial distributions, with the reactant n-butane
being mainly 13C2-n-butane. This indicates that the isomerization proceeded with an intermolecular rearrangement through a bimolecular
mechanism. In the temperature range of 493–523 K over these solid acids, the isotopic distributions in isobutane deviated from the binomial
distributions; the fractions of 13C2-isobutane were greater than those expected from the binomial distributions, showing that an intramolecular
(monomolecular) rearrangement became significant. Further, the product isobutane was found to consist of exclusively 13C2-isobutane over
these Pt-promoted catalysts in the presence of H2, demonstrating the operation of a monomolecular mechanism on the bifunctional catalysts.
© 2003 Elsevier B.V. All rights reserved.
Keywords: Isomerization of n-butane; Reaction mechanism; 1,4-13C-n-butane; Solid acid; Bifunctional catalyst; Field-ionization mass spectrometry
1. Introduction
12-tungstophosphoric acid, Cs2.5H0.5PW12O40, selectively
catalyzes the isomerization of n-butane even at 573 K [6,7].
WO3/ZrO2 was reported to be active for the reaction [8,9].
Besides these solid acids, noble metal, such as Pt-promoted
solid acids are promising catalysts in the presence of H2.
Pt-SO42−/ZrO2 is highly active for the isomerization of
n-butane [10,11] and has been utilized as a commercial
catalyst for isomerization of C5 and C6 hydrocarbons [12].
Pt-Cs2.5H0.5PW12O40 was reported to be highly selective
(>95%) in the presence of H2 [13,14]. Pt-WO3/ZrO2 also
showed a high selectivity for isomerization of n-butane,
although the activity is not high [15,16].
Since the selectivity to isobutane in the isomerization
of n-butane is closely related to the reaction mechanism,
understanding the mechanism is indispensable for devel-
opment of appropriate catalysts. It is generally accepted
that this reaction occurs through two possible pathways:
a monomolecular pathway with intramolecular rearrange-
ment, and a bimolecular pathway accompanying intermolec-
ular rearrangement [17]. One of the promising approaches
for elucidation of the reaction mechanism is utilization of
Skeletal isomerization of n-butane is of practical impor-
tance because the product isobutane is the main intermediate
for alkylation with butenes to form clean high-octane gaso-
line (C8-branched alkanes). The reaction of n-butane is thus
of particular interest, although among n-alkanes, n-butane
is much less reactive for the skeletal isomerization. Many
reports on this reaction have been published. Liquid acids
such as HF are active [1,2], but these have problems of
high toxicity, catalyst waste, use of large amounts of cat-
alyst, and difficulty in separation and recovery. Solid acids
are environmentally benign, because the above problems
are eliminated. The solid acids are required to have high
acid strength because of the low reactivity of n-butane.
Sulfated zirconia (SO42−/ZrO2) catalyzes this reaction at
low temperatures [3–5]. In addition, an acidic Cs salt of
∗
Corresponding author. Tel.: +81-11-706-4513;
fax: +81-11-706-4513.
E-mail address: oku@ees.hokudai.ac.jp (T. Okuhara).
1381-1169/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2003.08.021

146
T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
13C-labeled butanes as reactants. Analysis of the isotopic
distributions of the product butanes and the location of 13C
in the butanes gives helpful information on the reaction
mechanism. Thus far, mechanistic studies on the n-butane
isomerization have been done over SO42−/ZrO2 [18,19]
and H-mordenite [20] using 13C-labeled butanes. However,
the results are contradictory, and the mechanism remains a
point of controversy.
(0.108 mol dm−3, 34.6 cm3) was added to the solution un-
der vigorous stirring, forming a milky suspension. After
evacuation to remove water, the resulting solid was calcined
at 573 K in air for 2 h. The surface area of Pt-Cs2.5 was
80 m2 g−1
.
Sulfated zirconia (abbreviated as SZ) was prepared by
the method described in the literature [11,23]. Zr(OH)4 was
obtained from an aqueous solution of ZrCl2O·8H2O (Wako
Pure Chemicals Ind. Ltd.) by precipitation with an aqueous
ammonia at pH = 9, and the precipitate was filtered and
washed with water repeatedly until the filtrate was neutral.
The obtained solid was dried at 373 K for 24 h. A total
of 100 cm3 of H2SO4 (1 mol dm−3) was added to 9.9 g of
Zr(OH)4 under stirring, and the stirring was continued for
1 h. Then, the solid was filtered, dried at 373 K for 24 h,
and calcined at 893 K in air for 3 h. The surface area was
90 m2 g−1. Pt-sulfated zirconia (1.0 wt.%, abbreviated as
Pt-SZ) was prepared by impregnating SZ with the aqueous
solution of H2PtCl6 (0.041 mol dm−3). The obtained solid
was calcined at 573 K in air for 2 h, and its surface area was
According to Grain et al. [18], this reaction occurs through
the monomolecular mechanism at 523 K over SO42−/ZrO2
and Pt-SO42−/ZrO2. In contrast, Adeeva et al. [19] have im-
plied that it proceeds through the bimolecular mechanism at
353 K over SO42−/ZrO2. Matsuhashi et al. [21] presumed
that over SO42−/ZrO2, the monomolecular mechanism dom-
inates at the initial stage of reaction and the bimolecular
mechanism dominates at the latter stage. We preliminarily
reported that over Cs2.5H0.5PW12O40, this reaction takes
place in a parallel manner via both monomolecular and bi-
molecular mechanisms [22], and that the monomolecular
mechanism mainly operates for Pt-Cs2.5H0.5PW12O40 [23].
In the present study, we attempted to systematically
study the mechanism with typical solid acids and the
Pt-promoted catalysts by using 1,4-13C2-n-butane. We wish
to emphasize that the isotopic composition was analyzed
by field-ionization mass spectrometry (FI-MS). FI-MS is a
“soft” ionization technique in which a relatively small quan-
tity of internal energy is supplied to the molecule [24], so
that only the parent peak is detected [22,23]. From this, we
can quantitatively estimate the contribution of the reaction
pathways to discuss the contribution of each mechanism.
85 m2 g−1
.
Tungstena–zirconia (W/Zr ratio = 0.1, abbreviated as
WZ) was prepared by an impregnation method. Powder
of ammonium para-tungstate (NH4)10W12O41·5H2O was
first added to hot water (not dissolved instantly) and then
the suspension was stirred under ultrasonic radiation for
5 min to form the solution. Zr(OH)4 (10 g, Daiichi Ki-
genso Co.) that had been dried at 373 K overnight was
impregnated with an aqueous solution (0.016 mol dm−3) of
(NH4)10W12O41·5H2O at room temperature. The solid was
dried overnight at 373 K, and calcined at 1073 K in air for
3 h. Pt-WO3/ZrO2 (1.0 wt.%, abbreviated as Pt-WZ) was
prepared by incipient-wetness impregnation using the aque-
ous solution of H2PtCl6 by the same method as employed
for Pt-SZ. The obtained solid was calcined at 573 K in air
2. Experimental
2.1. Catalyst preparation
for 2 h, and its surface area was 70 m2 g−1
.
Cs2.5H0.5PW12O40 (abbreviated as Cs2.5) was pre-
pared according to the literature [25,26]. An appropriate
amount of the aqueous solution (0.10 mol dm−3) of Cs2CO3
(Merck, >99%) was added dropwise at a constant rate of
2.2. Materials
As a reactant, n-butane (Takachiho Chemical Ind. Co.
Ltd., Ultra pure) was used without further purification.
1,4-13C2-n-butane (13C: 99%) purchased from ISOTEC Inc.
was used without further purification.
about 1 cm3 min−1 to an aqueous solution (0.08 mol dm−3
,
20.0 cm3) of H3PW12O40 (Nippon Inorganic Color and
Chemicals Co.) at room temperature under vigorous stir-
ring. From the beginning of addition of Cs2CO3, very fine
particles (precipitates) were formed, making the solution
milky. After aging overnight at room temperature, water
was slowly removed by evaporation at 323 K. The resulting
white solid was ground into powder. The surface area was
found to be 120 m2 g−1 by the BET method.
2.3. Catalytic reaction
Skeletal isomerization of n-butane was performed at
393–623 K in a closed circulation system (200 cm3) made of
Pyrex glass with greaseless cocks equipped with an on-line
GC. After the catalyst was pretreated in vacuum at 573 K
for 2 h (for Cs2.5, WZ, Pt-Cs2.5, and Pt-SZ) or at 673 K
for 4 h (for SZ and Pt-SZ), 40 Torr (1 Torr = 133 Pa) of
n-butane (or 1,4-13C2-n-butane) was introduced. In the case
of Pt-promoted catalysts, 200 Torr of H2 was also added to
the reaction system. After the reaction, the reactant n-butane
and products including isobutane and C1–C8 hydrocarbons
Pt-Cs2.5H0.5PW12O40 (abbreviated as Pt-Cs2.5) was
also prepared according to the literature [13,14]. Amount
of Pt was adjusted to 1.0 wt%. An aqueous solution
(0.041 mol dm−3
,
12.1 cm3) of H2PtCl6 (Wako Pure
Chemical Ind. Ltd.) was added to the aqueous solution
(0.148 mol dm−3, 20 cm3) of H3PW12O40 at room tem-
perature. Subsequently, the aqueous solution of Cs2CO3

T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
147
were analyzed by an on-line GC (Hitachi, Gas Chromato-
graph 163) with a column of VZ-7 (3 mm × 6 m). In the
case of 1,4-13C2-n-butane, the reactant and product were
separated and analyzed by FI-MS (JEOL JNM-ECP400)
to determine 13C-distribution, and by 13C NMR (JEOL
EX-400) to determine the 13C-location in butanes.
3. Results
Fig. 1 shows the time courses of the skeletal isomeriza-
tion of n-butane over the solid acids at 523 K. In the initial
stage the reaction of n-butane proceeded rapidly on SZ, but
after about 30 min of reaction time the reaction became slow.
Among these catalysts, WZ was least active. Cs2.5 showed
a moderate activity. The deactivations for WZ and Cs2.5
were only slight. The selectivities to isobutane over SZ and
Cs2.5 were high in the initial stage, and decreased gradu-
ally with reaction time. In contrast, the selectivity over WZ
Fig. 2. Time courses of skeletal isomerization of n-butane over Pt-promo-
ted catalysts in the presence of H2 at 523 K. (᭿) Pt-Cs2.5H0.5PW12O40
,
(᭡) Pt-sulfated ZrO2, and (᭹) Pt-WO3/ZrO2. Reaction conditions:
n-butane 40 Torr; H2 200 Torr; catalyst weight 0.2, 0.05, and 0.5 g for
Pt-Cs2.5H0.5PW12O40, Pt-sulfated ZrO2, and Pt-WO3/ZrO2, respectively.
increased with reaction time and became nearly constant af-
ter 120 min. By-products on WZ in the initial stage were C2
and C3 hydrocarbons. Among these catalysts, Cs2.5 showed
the highest selectivity. At 180 min of the reaction time, the
selectivity was in the following order: Cs2.5 > WZ > SZ.
Fig. 2 presents the time courses of the skeletal isomeriza-
tion of n-butane over Pt-promoted catalysts in the presence
of H2 at 523 K. Pt-SZ was highly active, but considerable
deactivation was observed. The conversions for Pt-Cs2.5 and
Pt-WZ increased linearly with reaction time, indicating little
deactivation. Pt-SZ and Pt-Cs2.5 showed high selectivities
to isobutane (>90% at 300 min). The selectivity for Pt-WZ
was low because of formation of C1–C3 hydrocarbons by
the hydrogenolysis reaction.
Fig. 1. Time courses of skeletal isomerization of n-butane over typical
solid acids at 523 K. (᭿) Cs2.5H0.5PW12O40, (᭡) sulfated ZrO2, and
(᭹) WO3/ZrO2. Reaction conditions: n-butane 40 Torr; catalyst weight
0.5, 0.5, and 2.0 g for Cs2.5H0.5PW12O40, sulfated ZrO2, and WO3/ZrO2,
respectively.
Table 1 summarizes the catalytic activities and selectivi-
ties for the skeletal isomerization of n-butane. Comparison
Table 1
Catalytic activity and selectivity for skeletal isomerization of n-butanea
Catalyst
Rate ( ␮mol g−1 h−1
)
C1–C2
Selectivity (mol%)
C3
iso-C4
C5
C6–C8
WO3/ZrO2
Cs2.5H0.5PW12O40
SO42−/ZrO2
8.1
64.5
752.5
4.0
0.4
0.2
8.7
6.8
13.8
83.2
87.1
78.3
4.1
5.7
7.7
0.0
0.0
0.0
Pt-WO3/ZrO2
Pt-Cs2.5H0.5PW12O40
Pt-SO42−/ZrO2
14.1
57.9
2523.6
29.7
2.2
7.0
20.2
3.6
7.9
49.5
92.0
84.4
0.6
2.2
0.7
0.0
0.0
0.0
a
The reaction was performed at 523 K with n-butane 40 Torr. H2 200 Torr was added over Pt-promoted catalysts. The data was taken at about 10%
conversion.

148
T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
at ca.10% of the conversion at 523 K shows that the reaction
rate was higher over the Pt-promoted catalysts than over the
unpromoted catalysts, except for Cs2.5. With regard to se-
lectivity, Pt-promoted catalysts other than Pt-WZ are more
selective than the unpromoted solid acids. Pt-WZ yielded
large amounts of C1–C3 alkanes as by-products, because of
the hydrogenolysis reaction. In the case of the solid acids, C3
and C5 hydrocarbons were formed as by-products (Table 1),
indicating that dimerization-cracking (bimolecular mecha-
nism) proceeded over these catalysts, as will be discussed
below. In contrast, the C5 hydrocarbons were formed very
little on Pt-SZ and Pt-Cs2.5.
Fig. 3 shows the isotopic distributions of n-butane
and isobutane over Cs2.5 and WZ. Binomial patterns of
13C-isobutane (open rectangles in Fig. 3) were calculated
from the content ratio (r) of 13C to 12C, assuming bino-
mial distribution, 1:4r:6r2:4r3:r4, where r was determined
by optimization between the observed and calculated frac-
12
tions for x = 0, 1, 3, 4 in 13Cx
C
H10 of isobutane
4−x
[22,23]. For both catalysts, n-butane mainly consisted
of 13C2-n-butane (Fig. 3a). The isotopic distributions of
isobutane at 423 K on both catalysts were very close to
the binomial distributions (Fig. 3b). This shows that the
isomerization over these catalysts at 423 K proceeds via
an intermolecular process and that the rearrangement is
rapid, which accords with the previous results for SZ [22].
In contrast, at 523 K both the catalysts yielded mainly
13C2-isobutane (Fig. 3c), whose fractions were much
greater than those of the binomial ones. The patterns shown
in Fig. 3c demonstrate that the isomerization proceeds
through both the competitive pathways, intermolecular and
intramolecular.
Table 2 summarizes the isotopic distributions of isobu-
tane and n-butane. Table 2 also shows the ratios of observed
fraction of 13C2-isobutane to the calculated fraction. These
ratios were close to unity (1.3 for Cs2.5 and 1.3 for WZ)
at 423 K, and were nearly independent of the conversions,
e.g., 1.3 and 1.5 at the conversions of 11.0 and 27.5%, re-
spectively, on Cs2.5.
Fig. 3. 13C-distribution of n-butane and isobutane in the isomerization of
1,4-13C2-n-butane over (A) Cs2.5H0.5PW12O40 and (B) WO3/ZrO2. (a)
n-Butane at 423 K, (b) isobutane at 423 K, and (c) isobutane at 523 K.
The conversions were around 10%.
against the reaction temperature. The contribution of the
monomolecular pathway was estimated from the amounts
by which the 13C2-isobutane fraction exceeded that of the
calculated binomial distribution [22,23]. At low temper-
atures, the contributions of the monomolecular pathway
were very low (less than 12%) for all solid acids. A note-
worthy finding is that the contributions of monomolecular
pathway showed similar trends; that is, they increased with
increasing temperature. Within these temperature ranges,
the contributions of the monomolecular pathway were still
less than 50%.
In Fig. 4, the contribution of intramolecular pathway
(monomolecular mechanism) on these solid acids is plotted
Table 2
Selected values of observed isotopic distributions of isobutane and n-butane in the reaction of 1,4-13C2-n-butanea
Catalyst
Temperature
(K)
Conversion
(%)
Isobutane (%)
n-Butane (%)
13C0
13C1
13C2
13C3
13C4
13C0
13C1
13C2
13C3
13C4
Cs2.5H0.5PW12O40
423
423
523
11.0
27.5
12.8
13.0
5.8
6.6
21.9
22.7
13.8
41.9 (1.3)
46.9 (1.5)
62.6 (3.1)
19.3
20.1
14.9
3.9
4.5
2.1
0.7
1.5
1.8
2.8
5.0
2.9
94.3
89.7
93.1
2.2
3.2
2.1
0.0
0.6
0.1
WO3/ZrO2
Pt-Cs2.5H0.5PW12O40
423
423
10.2
10.1
6.8
2.8
23.7
4.3
43.2 (1.3)
90.6 (17.8)
22.0
2.1
4.3
0.2
0.6
1.7
2.3
2.5
95.7
93.7
1.4
2.1
0.0
0.0
Values in parentheses are the ratios of the observed fractions of 13C2 isobutane to the calculated fractions.
a
Reaction conditions; n-butane 40 Torr. H2 200 Torr was added over Pt-Cs2.5H0.5PW12O40. Isotopic distributions were estimated with field-ionization
mass spectrometry.

T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
149
Fig. 6. Contribution of the monomolecular pathway in skeletal isomeriza-
Fig. 4. Contribution of the monomolecular pathway in skeletal isomeriza-
tion of n-butane over solid acids. (᭿) Cs2.5H0.5PW12O40, (᭡) sulfated
ZrO2, and (᭹) WO3/ZrO2.
tion of n-butane over Pt-promoted catalysts. (᭿) Pt-Cs2.5H0.5PW12O40
(᭡) Pt-sulfated ZrO2, and (᭹) Pt-WO3/ZrO2.
,
were quite different from the binomial distributions; the
fractions of 13C2-butane were much greater than those
of the binomial distributions. These patterns indicate
that the isomerization on Pt-promoted catalysts proceeds
through the intramolecular pathway (the monomolecular
mechanism).
Fig. 5 shows the isotopic distributions of n-butane and
isobutane over Pt-Cs2.5 and Pt-WZ. It was confirmed that
n-butane consisted mainly of 13C2-butane over both the
catalysts (Fig. 5a). The isotopic distributions of isobutane
In Fig. 6, the contribution of the monomolecular mech-
anism over Pt-promoted catalyst is plotted against re-
action temperature. In contrast to the results on the
solid acids (Fig. 4), the contributions of monomolecular
mechanism were high at these temperatures over all the
Pt-promoted catalysts. Especially, the contributions ex-
ceeded 80% at low temperature. Contrary to the trend in
Fig. 4, the contributions tended to decrease with increasing
temperature.
Fig. 7 shows the effects of hydrogen pressure on the con-
tribution of the bimolecular mechanism, selectivity to C5 hy-
drocarbons, and rates of isomerization and hydrogenolysis
on Pt-WZ at 473 K. As is clear from the figure, the changes
in the contribution of bimolecular pathways and the selec-
tivity to C5 hydrocarbons are similar. As hydrogen pressure
increased, the rate for the isomerization decreased and that
for the hydrogenolysis increased.
In order to determine the location of 13C in n-butane
after the reaction, the 13C NMR was measured. Before
the reaction, a single peak at 13.8 ppm (methyl carbon)
was detected. In contrast, after the reaction (conversion
of n-butane = 12.4%) over Cs2.5 at 423 K, a new peak
at 24.9 ppm (methylene carbon) was observed, in addition
to the peak at 13.8 ppm. The relative peak intensities for
methyl and methylene carbon were 94.3 and 5.7%, respec-
tively. When n-butane containing 13C in natural abundance
was measured, the two peaks were detected with almost
the same intensity, indicating similar sensitivities of carbon
atoms for the NMR measurement.
Fig. 5. 13C-distribution of n-butane and isobutane in the isomerization of
1,4-13C2-n-butane over (A) Pt-Cs2.5H0.5PW12O40 and (B) Pt-WO3/ZrO2.
(a1) n-Butane at 423 K, (a2) n-butane at 473 K, (b1) isobutane at 423 K,
(b2) isobutane at 473 K, (c1) isobutane at 523 K, and (c2) isobutane at
573 K. The conversions were about 10%.

150
T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
4. Discussion
4.1. Proposed reaction mechanism
The possible rearrangement patterns of 13C in the isomer-
ization when 1,4-13C2-n-butane is used as reactant are sum-
marized in Scheme 1. In the monomolecular mechanism,
the isomerization would involve a protonated methylcyclo-
propane and a sequent primary carbenium ion as interme-
diates. If the reaction proceeds through the monomolecular
mechanism, 13C2-isobutane in the topoisomers of isobutane
would be produced exclusively (intramolecular rearrange-
ment), together with 1,3-13C2-n-butane as an isotopomer by
self-isomerization, when 13CH3–(CH2)2–13CH3 was used
as a reactant (Scheme 1A). Since the cracking of secondary
carbenium ion directly to C1–C3 hydrocarbons is unfavor-
able from the viewpoint of energy [17], the selectivity to
isobutane will be 100% in this mechanism. On the other
hand, the bimolecular mechanism (Scheme 1B) is possible
if butenes and sec-butyl carbenium ion are formed on the
catalyst surface to form octyl cation [17]. The ␤-scission
of octyl cation would yield C4 moieties, together with the
by-products of C3 and C5 alkanes. In this case, intermolec-
ular isotopic scrambling in product isobutane is expected,
as shown in Scheme 1B.
Fig. 7. Effects of H2 pressure on (᭺) contribution of the bimolecular
pathway, (᭹) selectivity to C5 hydrocarbons, and rate of (᭛) isomeriza-
tion, and (᭜) hydrogenolysis on Pt-WO3/ZrO2 at 473 K.
The amounts of 13C2-isobutane and 1,3-13C2-n-butane
produced were determined from FI-MS of isobutane and
13C NMR of n-butane, respectively. The data for about
10% conversion were used. In Fig. 8, the ratio of rate
constants (kc: carbon-shift reaction and ks: skeletal iso-
merization in Scheme 1A) for Cs2.5 is plotted against the
reaction temperature. This ratio corresponds to that of the
amount of 1,3-13C2-n-butane to 13C2-isobutane formed
from the monomolecular pathway. As shown in Fig. 8,
this ratio is about 10 at 393 K but decreases greatly with
increasing temperature, reaching about 0.7 at 523 K on
Cs2.5.
4.2. Skeletal isomerization of n-butane over solid acids
In the skeletal isomerization of n-butane over solid
acids, one of the critical steps is activation of n-butane
to sec-carbenium ion. Two processes are possible, that
is, protonation of n-butane, and dehydrogenation ((1) in
Scheme 2) or hydride abstraction by Lewis acid ((2) in
Scheme 2). These two steps are supposed to require high
acid strength for the acidic sites. Although the subse-
quent reaction pathways such as the rearrangement of
carbenium ions and the hydride transfer reaction to form
isobutane are also proposed to be the critical steps [13],
in any event, the acid strength must sensitively influence
the catalytic activity. The order of the initial activity,
SZ > Cs2.5 > WZ (Table 1), can be explained by the acid
strengths of these solid acids. TPD of NH3 showed that the
acid strength of these catalysts was in the order of SZ >
Cs2.5 > WZ [27,28] and that the acid amount was similar
[25,28].
The selectivity over solid acids is expected to be closely
related to the reaction mechanism. As was explained above,
the bimolecular pathway would readily cause the cracking
reaction, whereas the isomerization reaction would selec-
tively proceed under the monomolecular pathway. Fig. 4
clearly demonstrates that the reaction took place via similar
pathways over all these solid acids, that is, the bimolec-
ular pathway at low temperatures and parallel pathways
(mono and bimolecular pathways) at high temperatures.
In conformity with this, the by-products were C3 and C5
hydrocarbons (Table 1). As is well known, the bimolecular
Fig. 8. Ratio of the amount of 1,3-13C2-n-butane to that of 13C2-isobutane
over Cs2.5H0.5PW12O40
13C2-isobutane were estimated from FI-MS of isobutane and 13C NMR
.
The amounts of 1,3-13C2-n-butane and
of n-butane, respectively. The conversions were about 10%.

T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
151
Scheme 1. (A) Monomolecular pathway. (B) Bimolecular pathway.
pathway is favorable from the viewpoint of energy, because
this pathway proceeds via sec-carbenium ion. On the other
hand, the monomolecular pathway must path through a pri-
mary carbenium ion intermediate (Scheme 1). Therefore,
a reasonable conclusion is that the bimolecular pathway
was preferential at low temperatures over all these solid
acids. The increased contribution of monomolecular path-
way with increasing reaction temperature can be explained
by the large activation energy for the monomolecular
pathway [17]. An intriguing finding is that the contribu-
tion of the reaction pathway was essentially the same for
these solid acids, although acid strength differed greatly
[27].
Scheme 2.
4.3. Skeletal isomerization of Pt-promoted catalysts
(4)
(5)
+ H
Over the bifunctional catalysts (metal-promoted solid
acids), the bifunctional mechanism has been proposed
[29,30]. In this mechanism, first, n-butane is dehydro-
genated to butenes ((4) in Scheme 3), and then, the butenes
are protonated on acidic sites to form sec-carbenium ion
((5) in Scheme 3). The steps represented in (4) in Scheme 3
are in equilibrium. These steps are supported by the follow-
2
+H
Scheme 3.

152
T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
ing: (1) the reaction rate is independent of the Pt content,
and (2) the rate is inversely dependent on the hydrogen
pressure (Fig. 7) [13].
In the bifunctional mechanism, the activation of n-butane
is expected to be easier than in the acidic mechanism ((1)
and (2) in Scheme 2). Higher reaction rates were obtained
for the bifunctional catalysts (Table 1), which is probably
due to the bifunctional mechanism.
In the presence of H2 over the bifunctional catalysts,
the content of alkenes would be greatly suppressed, which
is unfavorable for the bimolecular mechanism. Consistent
with the above idea, the reaction over these bifunctional
catalysts proceeded exclusively via the monomolecular
pathway (Fig. 6). As the reaction temperature increased,
the contribution of the bimolecular mechanism increased.
This is probably due to the increase in the equilibrium
concentrations of butenes.
As shown in Fig. 7, the contribution of bimolecular path-
way and the selectivity to C5 hydrocarbons changed sim-
ilarly to the hydrogen pressure over Pt-WZ. This result is
consistent with the expectation from Scheme 1.
Fig. 9. Proposed model of the methyl cyclopropane intermediate on the
surface of the catalyst.
5. Conclusion
The reaction mechanism of the skeletal isomerization of
n-butane over solid acids such as Cs2.5H0.5PW12O40, sul-
fated ZrO2, and WO3/ZrO2 and their Pt-promoted catalysts
were investigated by using 1,4-13C2-n-butane as reactant.
The reaction took place via similar pathways independently
of the kind of solid acids, that is, via a bimolecular pathway
at low temperatures and via parallel (mono and bimolecu-
lar) pathways at high temperatures. On the other hand, the
monomolecular mechanism operated mainly the isomer-
ization reaction over Pt-promoted catalysts over the wide
range of the reaction temperature.
4.4. Rearrangement of carbon atom in protonated
cyclopropane intermediate
In the monomolecular mechanism for n-butane isomer-
ization, protonated methylcyclopropane intermediate has
been proposed [17]. The skeletal isomerization will be
brought about by the cleavage of the bond “a” in (3) in
Scheme 2. On the other hand, the cleavages of the other
bonds, “b” and “c” in (3) in Scheme 2, will cause the shift
of carbon, giving back to n-butane. The former must pass
through a primary carbenium ion, whereas the latter can pass
through the protonated methylcyclopropane sec-carbenium
ions (in Scheme 1A). The energy levels of these ions are
in the order of sec-carbenium ion > protonated methyl
cyclopropane > primary carbenium ion [17]. As a matter of
fact, the carbon-shift reaction took place more rapidly than
the skeletal isomerization of n-butane over H2SO4 [17],
which supports the above mechanism.
Acknowledgements
This work was partially supported by Core Research for
Evolutional Science and Technology (CREST) of Japan
Science and Technology Corporation (JST).
References
[1] G.A. Olah, G.K.S. Parakash, J. Sommer, Superacids, Wiley/
Interscience, New York, 1985.
[2] G.A. Olah, O. Farooq, A. Husain, N. Ding, N.J. Trivedi, J.A. Olah,
Catal. Lett. 10 (1991) 239.
As shown in Fig. 8, the carbon-shift reaction proceeded
more than 10 times as fast as the skeletal isomerization
over Cs2.5 at 423 K, which is consistent with the result
over H2SO4. However, at 523 K, the ratio in reaction
rates decreased to about 0.7. The difference in activation
energy between the protonated methylcyclopropane and
sec-carbenium ions obtained for Cs2.5 was estimated to be
13.2 kJ mol−1, which is much smaller than that for H2SO4
(60 kJ mol−1) [17]. This difference may be brought about
by the stabilization of ions through the adsorption on the
catalyst surface. Furthermore, about the fact that the skeletal
isomerization was slightly more rapid than the carbon-shift
reaction at 523 K, we speculate that the methylcyclopropane
intermediate orients specially in the manner of end-on [31]
shown in Fig. 9, which make the cleavage of the bond “a”
in (3) in Scheme 2 easier.
[3] H. Hino, K. Arata, J. Am. Chem. Soc. 101 (1979) 6439.
[4] K.T. Wan, C.B. Khouw, M.E. Davis, J. Catal. 158 (1996) 311.
[5] C.-Y. Hsn, C.R. Heimbuch, C.T. Armes, B.C. Gates, J. Chem. Soc.,
Chem. Commun. (1992) 1645.
[6] K. Na, T. Okuhara, M. Misono, J. Chem. Soc., Faraday Trans. 1 91
(1995) 367.
[7] N. Essayem, G. Coudurier, M. Fournier, J.C. Vedrine, Catal. Lett.
34 (1995) 223.
[8] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1987) 1259.
[9] J.C. Yori, C.L. Pieck, J.M. Parera, Appl. Catal. A 181 (1999) 5.
[10] M. Hino, K. Arata, Catal. Lett. 30 (1995) 25.
[11] K. Ebitani, J. Konishi, H. Hattori, J. Catal. 131 (1991) 257.
[12] N.A. Cusher, Handbook of Petroleum Refining Process,
McGraw-Hill, New York, 1996.
[13] K. Na, T. Okuhara, M. Misono, J. Catal. 170 (1997) 96.
[14] T. Okuhara, T. Yamada, K. Seki, K. Johkan, T. Nakato, Micropor.
Mesopor. Mater. 21 (1998) 637.
[15] J.C. Yori, J.M. Parera, Catal. Lett. 65 (2000) 205.

T. Echizen et al. / Journal of Molecular Catalysis A: Chemical 209 (2004) 145–153
153
[16] S. Eibl, R.E. Jentoft, B.C. Gates, H. Knözinger, Phys. Chem. Chem.
Phys. 2 (2000) 2565.
[25] T. Okuhara, T. Nishimura, H. Watanabe, M. Misono, J. Mol. Catal.
74 (1992) 247.
[17] B.C. Gates, Catalytic Chemistry, Wiley, New York, 1992.
[18] F. Grain, L. Seyfried, P. Girard, G. Marie, A. Abdulsamad, J. Sommer,
J. Catal. 151 (1995) 26.
[19] V. Adeeva, G.D. Lei, W.M.H. Sachtler, Catal. Lett. 33 (1995) 135.
[20] C. Bearez, F. Avendano, F. Chevalier, M. Guisnet, Bull. Soc. Chim.
Fr. 3 (1985) 346.
[26] S. Tatematsu, T. Hibi, T. Okuhara, M. Misono, Chem. Lett. (1984)
865.
[27] T. Okuhara, T. Nishimura, M. Misono, Stud. Surf. Sci. Catal. 101
(1996) 581.
[28] L. Li, Y. Yoshinaga, T. Okuhara, Phys. Chem. Chem. Phys. 1 (1999)
4913.
[21] H. Matsuhashi, H. Shibata, H. Nakamura, K. Arata, Appl. Catal. A
187 (1999) 99.
[29] G.L. Frischkorn, P.J. Kuchar, R.K. Olson, Energy Prog. 8 (1988)
154.
[22] T. Suzuki, T. Okuhara, Chem. Lett. (2000) 470.
[23] T. Suzuki, T. Okuhara, Catal. Lett. 72 (2001) 111.
[24] G.A. St. John, S.E. Buttrill Jr., M. Anbar, Organic chemistry of
coal, in: J.W. Larsen (Ed.), ACS Symposium Series 71, American
Chemical Society, Washington, DC, 1978, pp. 223–239.
[30] B.C. Gates, J.R. Katzer, G.C. Schuit, Chemistry of Catalytic Pro-
cesses, McGraw-Hill, New York, 1979.
[31] F. Chevalier, M. Guisnet, R. Maurel, in: G.C. Bond, P.B. Wells, F.C.
Tompkins (Eds.), Proceedings of the 6th International Congress on
Catalysts, London, 1976, p. 478.