Hydrogenolysis of n-hexane on Al2O3-supported Ir catalysts of various treatments

Article abstract of DOI:10.1006/jcat.1999.2621

A high-dispersion 1 wt % Ir/Al₂O₃ catalyst with 1 wt % Cl and its counterparts with < 0.2% Cl with or without sintering were tested in n-hexane reactions at 453-603 K. Dispersion and Cl content strongly influenced the activity of the catalyst, but the selectivity changes were not substantial. Hydrogenolysis, the main reaction with selectivities ≤90%, arose by superposition of a single and a multiple rupture, with higher temperatures and lower hydrogen pressures favoring the latter process.

Full text of DOI:10.1006/jcat.1999.2621

Journal of Catalysis 187, 486–492 (1999)
Article ID jcat.1999.2621, available online at http://www.idealibrary.com on
Hydrogenolysis of n-Hexane on Al O -Supported Ir Catalysts
2
3
of Various Treatments
�
,
�
�
�
,1
,1
A. Majest e´ , † S. Balcon, M. Gu e´ rin, C. Kappenstein, and Z. Pa a´ l†
�
LACCO, UMR-CNRS 6503, Catalyse en Chimie Organique, Facult e´ des Sciences, 86022 Poitiers Cedex, France; and †Institute of Isotope and Surface
Chemistry, CRC, Hungarian Academy of Sciences, Budapest, P.O. Box 77, H-1525 Hungary
Received April 28, 1999; revised June 28, 1999; accepted June 28, 1999
to give 2C3 on Ir black (11) (at 453 K), whereas an almost
A high-dispersion 1 wt% Ir/Al2O3 with 1 wt% Cl as well as its random rupture with some preference for terminal splitting
counterparts with <0.2% Cl with or without sintering were tested was reported at 400 K on Ir black prepared from IrO (4).
2
in n-hexane reactions between 453 and 603 K. The Cl-containing
catalyst was considerably less active in this temperature range, un-
favorable for bifunctional reactions. Hydrogenolysis was the main
reaction with selectivities up to 90%, along with minor amounts of
nondegradative products. Single hydrogenolysis with some prefer-
ence for splitting the molecule in the middle occurred at low temper-
atures. That was replaced gradually by multiple rupture at higher
temperatures. Increasing the hydrogen pressure at 513 K led to
higher hydrogenolysis selectivities up to 98%; at the same time, it
The ability of Ir to produce skeletal isomers and C5 cyclics
from alkane feed has been reported several times (1, 2, 6–8,
1
0–13).
Calcination of supported Ir catalysts seems to be a crucial
step in determining its further dispersion and morphology
4, 14, 15). Heating in air facilitated the formation of large
(
IrO2 crystallites, the reduction of which could result in Ir
metal of low dispersion (4). In the presence of Pt, highly
suppressed somewhat multiple fragmentation. With less hydrogen dispersed Pt–Ir clusters also appeared. The Cl content of
present (at lower H2 pressure or higher temperature), a “soft” tran- the support may have been critical for sintering: Huang
sition from single to multiple rupture occurred. The multiplicity et al. reported that IrO agglomeration did not occur above
2
of rupture was most pronounced on the larger crystallites of the
sintered sample, whereas a terminal rupture seemed to appear on
the Cl-free high-dispersion sample. Whenever single hydrogenoly-
sis prevailed, the fragment distribution showed a slight preference
a critical Cl content (15). The formation of large iridium
oxide crystallites as a precursor of the reduced metal was
regarded to be important to bring about “Ir-like” behavior
(
8).
Several studies have been published using appropriate
to propane formation.
�c 1999 Academic Press
Key Words: Ir/Al2O3; hydrogenolysis; n-hexane; Cl in alumina.
modelreactions, such asskeletalreactionsofn-hexane (nH)
for the characterization of the activity and selectivity pat-
tern of various catalysts. These studies included Pt (16–18),
Pt–Sn (18, 19), and Pt–Ir (20) catalysts. Whereas a “skele-
INTRODUCTION
Iridium is a rather active catalyst of alkane reactions (1, tal” reaction, such as hydrogenolytic opening of the cy-
). Pt–Ir catalysts have also gained an importance in naph- clopentane ring, showed a marked structure sensitivity on
2
tha reforming (2). To understand the action of such cata- Pt, it was found to be structure-insensitive on Ir (7). With
lysts, it is essential to study the catalytic properties of Ir respect to the possible importance of high-temperature cal-
alone or in comparison with those of Pt (1, 3–7). The hy- cination and the role of Cl in determining the final disper-
drogenolysis activity of iridium is much higher than that of sion, it seemed worth investigating the catalytic propensity
platinum (1, 4), although recent experiments (7, 8) have not of Ir/Al2O3 catalysts of various Cl contents and dispersions.
confirmed the large rate differences reported to be several These results are to be reported here. We found (20) that Ir
orders of magnitude (1, 4). Correspondingly, the isomer- could catalyze the isomerization of n-hexane, varied from
ization activity of Ir was, in turn, much lower than that of an earlier report (21) where isomerization was found with
Pt (3). In spite of the pronounced hydrogenolysis on Ir, methylpentane reactants only.
the complete rupture of the alkane molecules to methane
was not observed: the fragmentation could result in two or
just above two fragments per molecule split-up (9, 10). The
METHODS
main direction of splitting occurred in the middle position,
Catalyst BS100 was prepared by wet impregnation of a
Degussa Al2O3 (aluminium oxide C, �-alumina, BET sur-
face area, 95 m /g, pore volume, 0.7 cm /g, no micropores)
1
2
3
To whom correspondence should be addressed.
486
0021-9517/99 $30.00
Copyright �c 1999 by Academic Press
All rights of reproduction in any form reserved.

HYDROGENOLYSIS OF n-HEXANE
487
by H2IrCl6 � nH2O (Johnson Matthey) to give an Ir content calculated on the basis of moles of nH reacted rather than
of 1.0% . After drying at 393 K for 12 h, it was reduced considering the product moles formed. The hydrogenol-
in H2 flow for 2 h at 473 K and subsequently for 2 h at ysis reaction was characterized by (a) the � “fragmenta-
6
73 K (Dispersion: 48% , Ir, 1 wt% , Cl, 1.1 wt% , atomic ratio tion factor,” meaning the average number of fragments per
molecules fragmented (9, 13), and (b) the Mf “fission pa-
Cl/Ir � 6).
Catalyst BS131 was also prepared from sample BS100 in rameter,” relating the amount of larger fragments to that
an Ar flow containing 2% water and 1% H2 at 773 K for of methane (22). The parameter � permits us to differenti-
2
0 h (Dispersion: 40% (no sintering) but loss of Cl (less ate between “single” (� � 2) and “multiple” (� ꢀ 2) frag-
than 0.2 wt% , atomic ratio Cl/Ir � 1.))
mentation, whereas Mf distinguishes “terminal” (Mf = 1),
Catalyst BS105 was obtained by treating sample BS100 “random or internal” (Mf ꢀ 1), and “multiple” (Mf < 1) hy-
in an Ar flow containing 2% water at 773 K for 20 h (Dis- drogenolysis. Whenever hydrogenolysis was close to single
persion: 10% (sintering) and loss of Cl (less than 0.2 wt% , rupture, the ω “reactivity factor” (23) could also be used
atomic ratio Cl/Ir � 1)).
with some caution: it was calculated from the abundance of
Catalytic measurements were carried out in a closed-loop C3–C5 fragments, disregarding multiple rupture.
reactor setup described earlier (16, 18, 19). n-Hexane (nH,
Merck, GC grade) was used as model substance at a pres-
RESULTS
sure of 1.33 kPa with mixtures with hydrogen (8–64 kPa).
Temperature Dependence
The temperature varied between 453 and 603 K. More de-
tailed investigations were carried out at 513 and 603 K.
The conversion values were ca. 1% at 453 K, increasing
The amount of catalyst was selected in a way that they to above 5% at 603 K (Table 1). These corresponded to
should contain approximately the same number of exposed turnover frequency (TOF) values of 10–190. The lowest
17
metal atoms, namely, 3.2–3.4 � 10 Irsurf.. This way, 23 mg TOF values were obtained on the sample BS100. The other
of BS100, 27 mg of BS105, and 102 mg of BS131 were used. two samples, BS105 and BS131, showed higher values, by
Samples were taken after 5 min of reaction time and ana- about 30–40% , being rather close to each other.
lyzed by gas chromatography (16, 18). The catalysts were
Table 1 also shows selectivity values. Hydrogenolysis al-
regenerated by an oxygen–hydrogen treatment at reaction ways predominated, its selectivity being close to 90% in
temperature (16, 18). Conversions and selectivities were most cases. Of nondegradative products, the selectivity of
TABLE 1
Selectivities of n-Hexane Reactions on Various Ir Catalysts as a Function of Temperature
Selectivity (% )
TOF (h 1)
�
T (K)
Conv. (% )
<C6
Isomers
MCP
Benzene
Hexenes
(
A) Catalyst BS100
a
453
483
513
543
573
603
0.40
1.06
1.80
2.35
3.90
4.87
� 70
Sum C6: � 30
11
29
49
64
106
133
81.8
84.1
86.1
87.0
89.3
6.7
5.5
3.1
1.8
1.0
9.4
6.2
4.8
3.0
1.7
2.1
4.2
5.7
7.7
6.7
0
0
0.3
0.5
1.3
(
(
B) Catalyst BS131
4
4
5
5
5
53
83
13
43
73
1.17
1.85
2.10
3.45
4.70
88.1
87.8
86.9
87.8
88.2
4.6
5.8
4.7
3.3
1.8
7.3
5.1
5.1
3.3
2.4
0
0
0
0
0.1
0.3
34
54
62
102
139
1.3
3.3
5.5
7.3
C) Catalyst BS105
4
4
5
5
5
6
53
83
13
43
73
03
1.10
1.90
2.80
3.70
4.75
6.85
87.5
90.1
90.0
90.0
89.6
89.2
4.3
4.1
3.6
2.8
1.7
1.4
8.2
5.0
3.9
3.2
2.7
1.8
0
0
0
0
0.1
0.3
0.9
31
53
78
106
126
191
0.8
2.4
3.9
5.7
6.7
Note. p(nC6) = 1.33 kPa, p(H2) = 16 kPa; sampling time, 5 min.
a
Uncertain results due to product separation problems in the C6 range.

´
MAJESTE ET AL.
4
88
skeletal isomers and methylcyclopentane (MCP) decreased
and that of benzene increased at higher temperature. The
three samples showed no conspicuous differences. Benzene
appeared at lower temperatures than that reported for an
Ir film (24) (<500 vs 600 K); this film, in turn, showed no
isomerization activity.
TABLE 2
Characteristics of n-Hexane Hydrogenolysis on Ir Catalysts
as a Function of Temperature
Characteristic value
T (K)
�
Mf
ω1
ω2
ω3
Figure 1 illustrates that although the hydrogenolysis se-
lectivity was almost constant, the fragment distribution
showed conspicuous changes as a function of the reaction
temperature. Hydrogenolysis was almost random at 453 K.
Higher temperatures increased gradually the amount of
methane until it reached about 90% of all fragments.
This reaction showed somewhat more marked differences
on various samples: catalyst BS131 produced the lowest
amount of methane and the abundance of ethane was not
too low, even at 603 K.
(
A) Catalyst BS100
4
4
5
5
53
83
13
43
2.43
2.52
2.85
3.44
4.02
5.96
6.22
4.40
3.02
2.00
1.21
0.55
0.72
0.94
0.98
0.75
0.82
0.86
1.53
1.24
1.16a
573
603
(B) Catalyst BS131
453
483
513
543
573
2.25
2.48
2.88
3.31
3.90
8.80
4.88
3.14
2.30
1.45
0.69
0.84
0.83
0.80
0.89
0.93
1.52
1.27
1.24a
(
C) Catalyst BS105
453
483
513
543
573
603
2.29
2.53
2.94
3.46
4.08
4.70
7.73
4.83
3.14
2.07
1.23
0.60
0.62
0.92
0.94
0.68
0.75
0.85
1.72
1.33
1.22a
Note. Reaction conditions: p(nC6) = 1.33 kPa, p(H2) = 16 kPa; sam-
pling time, 5 min. The values of ω were calculated from the corresponding
larger fragments, ω1 from pentane, ω2 from butane, and ω3 from propane.
a
Values for ω1, ω2, and ω3 were equal to 0.84, 0.94, and 1.22, respectively,
on Ir films at 510 K (24).
The tendency of low temperatures favoring single hy-
drogenolysis was confirmed by the specific parameters for
hydrogenolysis (Table 2). At low temperatures the values
of � were just slightly above 2, pointing to a prevailing sin-
gle hydrogenolysis. The actual values of �ꢀ 2 indicate the
superposition of another, multiple hydrogenolysis to sin-
gle rupture. The continuous increase of � (Table 2) shows
that this multiple rupture took over at higher tempera-
tures. This agrees well with results reported for an Ir film
(
24). The decreasing values of Mf below unity indicate the
same trend: no terminal hydrogenolysis occurred. Values
of the ω reactivity factor are given for the lowest tempera-
tures, where � values were below 3; they show an enhanced
rupture in the middle of the molecule giving two propane
fragments. At 513 K, they are in rather good agreement with
the hydrogenolysis pattern reported for an Ir film at 510 K
(
24). The differences between the reactivities of various
C–C bonds decreased as multiple hydrogenolysis became
more and more important.
Hydrogen Pressure Effects
Curves with a maximum have been found to be character-
istic of yields of various products as a function of hydrogen
pressure (16, 17, 25, 26). Earlier studies reporting negative
FIG. 1. Fragment distribution as a function of temperature, normal-
ized to 6 (fragments) = 100% ; p(nC6) = 1.33 kPa, p(H2) = 16 kPa; sam-
pling time, 5 min.

HYDROGENOLYSIS OF n-HEXANE
489
tion temperatures (Table 4). A basically different behavior
is seen at these two temperatures.
At 513 K, the average number of fragments hardly ex-
ceeded the value of 2 (characteristic of single hydrogenol-
ysis), agreeing well with Mf values much higher than unity,
indicating random hydrogenolysis. Increasing the hydrogen
pressure shifted the splitting toward single hydrogenolysis
over all three samples. The random character of fragmenta-
tion (the value ofMf) also increased. It waspossible to calcu-
late also the ω reactivity factors for catalysts BS100. These
values (Fig. 3) indicate a shift of the preferential position
of bond rupture toward the middle C–C bond at higher hy-
drogen pressures. This was analogous to what was reported
earlier for Pt black (27).
At 603 K, the � fragmentation factor indicated multi-
ple hydrogenolysis. The hydrogen response of the samples
was different. Increasing p(H2) suppressed multiple split-
ting on the sample BS100 (containing Cl and having the
highest dispersion), even at this higher temperature. The
values of Mf showed more clearly the transition of multi-
ple to (commencing) single rupture from the lower to the
higher hydrogen pressure range. The fragmentation was
more pronounced with BS105 and BS131 where increasing
p(H2) increased its multiplicity. Splitting on BS105 (large
FIG. 2. Overall turnover frequency (TOF) values for three catalysts,
as a function of hydrogen pressure, p(H2), at T = 513 K, p(nC6) = 1.33 kPa;
sampling time, 5 min.
hydrogen exponents (1, 2) may have been carried out in the Ir crystallites, no Cl) gave almost entirely methane (� �
hydrogen pressure range higher than that corresponding to 6). The value of Mf being below unity confirmed the true
the maximum rate. A maximum hydrogenolysis rate has multiple character.
been found on Ir black at around 573 K (10). Higher hy-
drogen pressures were found recently to promote n-hexane
transformations with both Ir and Pt–Ir/Al2O3 catalysts be-
tween 513 and 603 K (20).
Figure 2 depicts TOF values as a function of the hydro-
TABLE 3
Selectivities of n-Hexane Reactions on Various Ir Catalysts
as a Function of Hydrogen Pressure
gen pressure at 513 K. These were lowest with the “origi-
Selectivity (% )
nal” sample BS100, exhibiting a maximum. Catalyst BS105
sintered, less Cl) was, in turn, most active, while the high-
p(H2)
a
(
kPa)
Conv. (% )
<C6
Isomers
Benzene
Unsat. + MCP
(
dispersion sample BS131 had an intermediate activity but
was closer to BS105. The hydrogen pressure monotonically
increased the intrinsic activity of these two latter catalysts.
Selectivity values are shown in Table 3 for all three cata-
lysts at a selected temperature of 513 K where the conver-
sion values were still rather low (2–8% ). Even at this not too
high temperature, 84–98% of n-hexane transformed into
fragments. The overall reactions of n-hexane were similar
for BS105 and BS131, both increasing, as a rule, at higher
hydrogen pressures. The maximum appeared also in hy-
drogenolysis selectivity on BS100 parallel to the maximum
in TOF. The selectivity of nondegradative reactions showed
a different p(H2) response. Their amount was nearly con-
stant on BS100—except for unsaturated products diminish-
ing at higher p(H2)—where Cl could, in principle, partici-
pate in the reactions. Higher hydrogen pressure suppressed
allC6 productsin a rather uniform wayon BS105and BS131,
possessing only metallic centers.
(
A) Catalyst BS100
8
16
2
8
4
1.3
1.6
2.0
2.6
1.5
84.1
84.1
87.0
90.8
87.5
3.2
4.8
5.1
4.2
4.2
8.5
7.9
6.0
4.2
7.4
4.2
3.2
1.9
0.8
0.9
3
4
6
(
B) Catalyst BS131
8
6
2.2
2.3
3.3
5.3
5.7
86.7
87.5
92.9
95.3
96.6
3.1
4.6
3.0
2.6
1.8
5.5
4.9
3.0
1.8
1.6
4.7
3.0
1.1
0.3
0
1
32
48
64
(
C) Catalyst BS105
8
6
2
8
2.6
3.0
5.1
6.8
8.4
90.2
91.0
95.6
97.5
98.0
2.0
3.0
1.9
1.5
1.1
4.6
3.7
1.8
1.3
0.9
3.2
2.3
0.7
0.2
0
1
3
4
64
Note. Reaction conditions: p(nC6) = 1.33 kPa, T = 513 K; sampling
time, 5 min. No hexenes were observed.
The hydrogen pressure response of two parameters for
hydrogenolysis, � and Mf, has been calculated for two reac-
a
Including hexenes, methylcyclopentane, and methylcyclopentene,
observed in some cases.

´
MAJESTE ET AL.
4
90
TABLE 4
idation and sintering of Ir (15). The customary preparation
of unsupported (5, 9, 10) and supported Ir catalysts (5–7,
2, 15) used its Cl-containing salt, except for its reduction
Characteristics of n-Hexane Hydrogenolysis on Ir Catalysts
1
as a Function of Hydrogen Pressure
from IrO2 (4) or the vacuum-deposited film. Acidification
of the alumina support of Ir by HCl enhanced the bifunc-
tional character (5). However, bifunctional reactions would
start at higher temperatures than those used here (5, 6) and
the oxidative treatment also required higher temperatures
to induce sintering (14) than that applied during regenera-
tion. The isomer selectivity in bifunctional isomerization at
Characteristic value
p(H2) (kPa)
� (513 K)
Mf (513 K)
� (603 K)
Mf (603 K)
(
(
(
A) Catalyst BS100
8
6
2
8
4
3.2
2.9
2.7
2.6
2.7
1.8
2.5
3.5
4.0
3.6
5.0
4.6
4.2
3.9
3.8
0.35
0.6
0.9
1.4
1.6
1
3
4
6
6
50 K could reach values of � 50% (5) as opposed to the few
percent reported here (Tables 1 and 3). So we can assume
that the reactions in our case could be attributed mainly
to Ir metallic sites, in agreement with the reports of Gault
et al. (3b, 28). n-Hexane contains no tertiary C atom; hence,
the “iso-unit” mode of reaction (3a) cannot be valid here.
Its ability to isomerize does not exclude the “1,3 bond shift”
mechanism reported with methylpentanes (3b, 21) but it is
clear now that this is not the only and exclusive method for
skeletal alkane isomerization on Ir catalysts.
B) Catalyst BS131
8
6
2
8
4
3.2
2.9
2.8
2.7
2.7
2.3
3.2
3.8
4.9
5.4
4.4
4.4
4.5
4.9
5.0
0.8
1.0
1.1
0.85
0.8
1
3
4
6
C) Catalyst BS105
8
6
2
8
4
3.6
3.1
3.0
2.9
2.9
2
4.8
4.7
4.8
5.7
5.7
0.45
0.65
0.7
0.2
0.2
1
3
4
6
2.9
3.6
4.2
4.4
The intrinsic activity of BS100 was the lowest at each hy-
drogen pressure investigated (Fig. 2). The presence of Cl
seemed to hamper the overall reaction. This effect was not
very remarkable at low p(H2) but rather significant at the
highest end of the hydrogen pressure range. The value of
Note. Reaction conditions: p(nC6) = 1.33 kPa, T = 513 K; sampling
time, 5 min.
�
1
� 1
TOF = 29 h at 483 K is not far from the TOF = 6.5 h
at 478 K reported on Rh/SiO2 for ethane hydrogenolysis,
which should be a more demanding reaction (2). Since hy-
drogenolysis was the main reaction on all catalysts, this
means that more H2 accelerated hydrogenolysis, at least at
the selected temperature, 513 K, as opposed to the nega-
tive H exponent between 450 and 480 K (1, 2, 4). The dif-
ferences in hydrogenolysis selectivities of various samples
were minor (Table 1). More Cl also reportedly decreased
hydrogenolysis selectivity of Pt (30)—an analogous phe-
nomenon may have appeared with the catalyst BS100. The
accumulation of Cl on the Ir–Al2O3 perimeter is not ex-
cluded here. Similar experiments with Pt/Al2O3 reported
DISCUSSION
The reactions on alumina-supported Ir catalysts were re-
ported to take place in a bifunctional way (2, 12). The pres-
ence of Cl may be important for facilitating the creation of
acid sites (5), apart from its reported role in controlling ox-
n+
Pt–Cl–O entities in the presence of Cl, giving rise to Pt
ions at these sites and also influencing redispersion (29).
The absence of Cl may have been the reason why the two
catalysts BS105 and BS131 with low Cl content showed
closer intrinsic activities and hydrogenolysis selectivities
at each temperature and hydrogen pressure. The highest
selectivity observed on the most active, sintered sample,
BS105, is in agreement with the generally accepted mech-
anism of hydrogenolysis requiring ensembles of several
metal atoms (2, 31). Its larger crystallites must have been
favorable for multiple hydrogenolysis (highest � values,
Table 4).
Activities as well as hydrogenolysis selectivity values in-
creased monotonically at higher temperatures (Table 1).
The slight decrease of hydrogenolysis selectivity at higher
temperature (12) did not appear here. At the same time,
increasing the hydrogen pressure had a similar but not very
FIG. 3. “Reactivity factors,” ω, for splitting n-hexane in different di-
rections on catalyst BS100, as a function of hydrogen pressure, p(H2), at
T = 513 K, p(nC6) = 1.33 kPa; sampling time, 5 minutes.

HYDROGENOLYSIS OF n-HEXANE
491
marked effect (Table 3). Increasing temperatures should (6% Pt/SiO2) were higher at 6-fold hydrogen excess, but
have decreased the hydrogen coverage at constant H2 pres- the Mf values were as low as 2–3 as opposed to Mf � 8 at
sure. The hydrogen coverage should have increased at a 48-fold hydrogen excess (16). Some authors regarded the
higher p(H2) values. However, both increasing temperature formation of “Ir–C” ensembles less likely (5, 6, 12). How-
and hydrogen pressure increased hydrogenolysis activity. ever, S a´ rk a´ ny attributed the isomerization ability of Ir to
A closer inspection of hydrogenolysis characteristics seems some adsorbed hydrocarbons also influencing the method
to resolve this apparent disagreement. The results indicate of reactive adsorption of the alkane feed (33). Ponec (2, 12,
two more or less independent fragmentation processes su- 34) reported higher hydrogenolysis selectivity at the low-
perimposed on each other: single and multiple hydrogenol- and the high-temperature ends of the temperature range.
ysis. The former is favored at higher hydrogen coverages: Such a marked temperature effect was not observed here.
at lower temperatures (Fig. 1, Table 2) and higher hydro-
To sum up, dispersion and Cl content strongly influence
gen pressures (Table 4). As the amount of available surface the activity of Ir/Al2O3 catalysts but the selectivity changes
hydrogen decreased, more and more molecules suffered a are not profound. The overwhelming hydrogenolysis arises
total disruption in addition to the less and less important by superposition of a “single” and a “multiple” rupture,
single hydrogenolysis (Tables 2 and 4). The competition of higher temperatures and lower hydrogen pressures favor-
the two processes is seen as a function of hydrogen pressure ing the latter process.
at 513 K, whereas single splitting is no more important at
6
03 K (Table 4). The continuous decrease of Mf values with
ACKNOWLEDGMENTS
the temperature indicates a “real” multiple rupture, i.e., the
disruption of the reactant into several fragments during one
A.M. thanks the French Ministry of Foreign Affairs for supporting her
sojourn on the surface rather than “slicing off” end methyl practice semester in Budapest during which the experiments were done.
S.B. thanks the French Space Agency, CNES (Centre National d’Etudes
Spatiales) for financial help. Z.P. is grateful to the University of Poitiers
for inviting him as a guest professor, rendering the completion of the
manuscript possible during this stay.
groups step by step. The value of Mf decreased from 17
to 8 from 450 to 550 K on Ir/Al2O3 (32). These values are
somewhat higher than those reported here. With the contin-
uous decrease of Mf from � 8 to below 1 (Table 2), its value
necessarily will reach and cross unity at a certain temper-
ature. This, however, may be an intermediate stage rather
REFERENCES
than “true” terminal splitting. One possible interpretation
1. (a) Sinfelt, J. H., Adv. Catal. 23, 91 (1973); (b) Sinfelt, J. H., and Yates,
of these results may be the increasing role of carbene-type
intermediates of multiple rupture with increasing temper-
ature, the formation of which reportedly has a high proba-
bility on Ir (3b, 28). This is in agreement with the decreas-
ing steady-state hydrogen coverage at higher temperatures
and also with increasing multiplicity at low p(H2) values
D. J. C., J. Catal. 8, 82 (1967).
2. Ponec, V., and Bond, G. C., “Catalysis by Metals and Alloys, Studies
Surface Science and Catalysis,” Vol. 95. Elsevier, Amsterdam, 1995.
. (a) Foger, K., and Anderson, J. R., J. Catal. 59, 325 (1979); (b) Gault,
F. G., Amir-Ebrahimi, V., Garin, F., Parayre, P., and Weisang, F., Bull.
Soc. Chim. Belg. 88, 475 (1979).
3
4
. Carter, J. L., Cusumano, J. A., and Sinfelt, J. H., J. Catal. 20, 223 (1971).
(
Tables 2 and 4).
The importance of the two types of hydrogenolysis was
5. Rasser, J. C., Beindorff, W. H., and Scholten, J. J. F., J. Catal. 59, 211
1979).
. McVicker, G. B., Collins, P. J., and Ziemiak, J. J., J. Catal. 74, 156
1982).
(
6
different with the three catalysts. The single-rupture charac-
ter was preserved on BS131 in the widest temperature range
(
(
7. Barbier, J., and Marecot, P., Nouv. J. Chim. 5, 393 (1981).
. Dees, M. J., and Ponec, V., J. Catal. 115, 347 (1989).
Fig. 1, Table 2), where the values of � were closest to those
8
observed on an Ir film (24). At 603 K, in turn, complete
methanation was almost reached on the sintered BS105,
its larger crystallites being favorable for multiple rupture:
9. Pa a´ l, Z., and T e´ t e´ nyi, P., React. Kinet. Catal. Lett. 7, 39 (1977).
10. S a´ rk a´ ny, A., Matusek, K., and T e´ t e´ nyi, P., J. Chem. Soc., FaradayTrans.
1
73, 1699 (1977).
1
1
1. Pa a´ l, Z., and T e´ t e´ nyi, P., React. Kinet. Catal. Lett. 12, 131 (1979).
2. Van Senden, J. G., Van Broekhoven, E. H., Vreesman, C. T. J., and
Ponec, V., J. Catal. 87, 468 (1984).
�
approached 6 and the Mf values were lowest at almost
each hydrogen pressure (Table 4). The position of single
rupture was also different with the three samples. A rather 13. Pa a´ l, Z., and T e´ t e´ nyi, P., Nature 267, 234 (1977).
1
1
4. Huang, Y.-J., and Fung, S. C., J. Catal. 118, 192 (1989).
5. Huang, Y.-J., Fung, S. C., Gates, W. E., and McVicker, G. B., J. Catal.
even distribution of fragments was reported with Ir reduced
from IrO2 (4). This was approached at a higher reaction
temperature (Table 2). With single rupture becoming more
and more important, splitting in the middle of the molecule
131, 378 (1991).
1
6. Pa a´ l, Z., Groeneweg, H., and Pa a´ l-Luk a´ cs, J., J. Chem. Soc., Faraday
Trans. 86, 3159 (1990).
became favored (Table 2, Fig. 3), in agreement with the re- 17. Pa a´ l, Z., Catal. Today 12, 297 (1992).
1
8. Pa a´ l, Z., Gyo˝ry, A., Uszkurat, I., Olivier, S., Gu e´ rin, M., Kappenstein,
sults obtained with n-hexane on Ir black (9), and Ir/SiO2
32). The internal splitting of n-pentane was also enhanced
at higher H2 excess (33). Terminal hydrogenolysis was at-
tributed to “Pt–C” ensembles (34). Indeed, hydrogenolysis
selectivities at the same conversion level on EUROPT-1
C., J. Catal. 168, 164 (1997).
(
1
9. (a) Kappenstein, C., Saouabe, M., Gu e´ rin, M., Marecot, P., Uszkurat,
I., and Pa a´ l, Z., Catal. Lett. 31, 9 (1995); (b) Kappenstein, C., Gu e´ rin,
M., L a´ z a´ r, K., Matusek, K., and Pa a´ l, Z., J. Chem. Soc., Faraday Trans.
94, 2463 (1998).

´
4
92
MAJESTE ET AL.
2
2
2
0. Charron, A., Kappenstein, C., Gu e´ rin, M., and Pa a´ l, Z., Phys. Chem. 28. Gault, F. G., Adv. Catal. 30, 1 (1981).
Chem. Phys. 3817 (1999).
29. Lieske, H., Lietz, G., Spindler, H., and V o¨ lter, J., J. Catal. 81, 8
1. Weisang, F., and Gault, F. G., J. Chem. Soc., Chem. Commun. 519
(1983).
(
1979).
2. Van Schaik, J. R. H., Dessing, R. P., and Ponec, V., J. Catal. 38, 273
1975).
30. Hayek, K., Kramer, R., and Pa a´ l, Z., Appl. Catal. A 162, 1
(1997).
(
31. Anderson, J. R., and Avery, N. R., J. Catal. 5, 446 (1966).
32. den Hartog, A. J., Holderbusch, M., Rappel, E., and Ponec, V., “Pro-
ceedings, 9th International Congress on Catalysis, Calgary, 1988”
(M. J. Phillips and M. Ternan, Eds.). Chem. Institute of Canada, Ot-
tawa, 1988.
2
2
2
2
2
3. Leclercq, G., Leclercq, L., and Maurel, R., J. Catal. 50, 87 (1977).
4. Plunkett, F. J., and Clarke, J. K. A., J. Catal. 35, 330 (1974).
5. Bond, G. C., and Slaa, J. C., J. Mol. Catal. 98, 81 (1995).
6. Bond, G. C., and Cunningham, R. H., J. Catal. 166, 172 (1997).
7. Zimmer, H., Dobrovolszky, M., T e´ t e´ nyi, P., and Pa a´ l, Z., J. Phys. Chem. 33. S a´ rk a´ ny, A., J. Chem. Soc., Faraday Trans. 1 85, 1511 (1989).
0, 4758 (1986). 34. Ponec, V., Adv. Catal. 32, 149 (1983).
9