New insights into the reaction mechanism and the rate-determining step of n-butane hydroisomerisation over reduced MoO3 catalysts

Article abstract of DOI:10.1039/b304978g

The comparison of the proportions of the minor products formed during the hydroisomerisation of n-butane over reduced MoO₃ with thermodynamic data demonstrates that a bifunctional mechanism operates and that the rate-determining step is the isomerisation of the linear butene intermediates to iso-butene.

Full text of DOI:10.1039/b304978g

New insights into the reaction mechanism and the rate-determining step
of n-butane hydroisomerisation over reduced MoO catalysts†
3
Frederic C. Meunier*
School of Chemistry, Queen’s University of Belfast, Belfast, Northern Ireland, UK.
E-mail: f.meunier@qub.ac.uk; Fax: +44 28 9038 2117; Tel: +44 28 90274420
Received (in Cambridge, UK) 2nd May 2003, Accepted 23rd May 2003
First published as an Advance Article on the web 19th June 2003
The comparison of the proportions of the minor products
yield of the main reaction product, i.e. iso-butane (right y-axis),
follows a sigmoid-shape curve, similarly to the data based on
other hydrocarbons.3 This behaviour can be explained by the
time needed to form the reactive phase(s) from the reduction of
formed during the hydroisomerisation of n-butane over
,10
3
reduced MoO with thermodynamic data demonstrates that
a bifunctional mechanism operates and that the rate-
determining step is the isomerisation of the linear butene
intermediates to iso-butene.
7
6
3 x y 2
MoO (e.g. an oxycarbide MoO C phase or MoO ) and/or the
11
increase in the surface area of the catalyst. The steady-state
value of the iso-butane yield (ca. 10%) was attained within 24
h under the experimental conditions used and corresponded to
an iso-butane selectivity of ca. 90%. The main competitive
The need to tackle global warming has renewed the interest in
alkane reforming for the production of more efficient and
cleaner fuels. Catalysts based on molybdena reduced at mild
reactions were cracking (i.e. C
tion (i.e. C + C ) and dehydrogenation (H
selectivity of ca. 8, 1 and 1%, respectively.
3
+ C
1
and 2 C
2
), disproportiona-
+ butenes), with a
temperatures, e.g. 623 K, are active for the hydroisomerisation
3
5
2
–5
4 7
of C –C
alkanes.1 These materials can be more selective than
zeolite-supported platinum (in particular at high conversions),
are more resistant to sulfur and nitrogen poisoning and do not
catalyse the formation of significant levels of aromatics.3
The yields of the butene products (left y-axis) during the
activation process are also reported in Fig. 1. Note that the total
concentration of butenes is only about one hundredth that of iso-
butane once steady-state conditions were reached. The concen-
tration ratios of reactant and products at various times are given
in Table 1. For the sake of clarity, the estimated error on the
ratios (typically 10–15%) is only reported for the steady-state
values. Interestingly, linear butenes were the first products to
appear, after one hour on stream. After about three hours (not
shown), but-1-ene was in thermodynamic equilibrium with n-
butane. As the three linear butenes were also in equilibrium at
–5
3
However, industrial applications based on reduced MoO are
not yet viable, partly due to the low intrinsic activity of these
5
materials. Further catalyst development has been hampered by
the structural complexity of the material and by contradictory
statements regarding the associated mechanism of reaction.
Katrib et al.6 and Matsuda et al.5 have proposed that a
traditional bifunctional mechanism operates, whereas Bouchy
7
et al. have proposed a metallacyclobutane intermediate-based
mechanism. In the case of the isomerisation of C
5
–C
7
alkanes
4
this point, all the linear C compounds were then equilibrated.
over reduced MoO
3
, Matsuda et al.5 proposed that the
Iso-butene was observed later than the linear butenes, i.e.
after two hours, and the ratio iso-butene/but-1-ene always
remained significantly below the corresponding thermody-
namic value (see Table 1). The formation of iso-butene showed
a maximum at ca. five hours, at which time the concentration of
iso-butane started to increase sharply. The steady-state concen-
trations of iso-butane and iso-butene remained significantly
below the thermodynamic limits associated with the isomerisa-
tion to the corresponding linear molecule (see values in bold
font in Table 1). However, it is important to stress that the
steady-state iso-butane/iso-butene ratio was that predicted by
the thermodynamic calculation.
dehydrogenation/hydrogenation process was the rate-determin-
ing step. The comparison of basic thermodynamic data with the
proportions of reaction products, even at the trace level, has
proven a valuable tool in unravelling reaction mechanisms.8
The data reported in the present article give a strong indication
on the nature of the alkane isomerisation mechanism taking
,9
3
place over reduced MoO samples and, moreover, suggest how
the activity of these catalysts could be greatly improved.‡
The experimental details are available on-line as Electronic
Supplementary Information.† The gradual increase of the
3
activity of MoO in n-butane conversion is shown in Fig. 1. The
These results clearly show that the first catalytic function
†
Electronic supplementary information (ESI) available: experimental
3
developed over the reduced and/or carburised MoO is a metal-
like function, which allows dehydrogenation/hydrogenation.
details. See http://www.rsc.org/suppdata/cc/b3/b304978g/
Fig. 1 Iso-butane and C
and (e) iso-butene.
4
-alkene yields during the reductive activation of MoO
3
at 623 K. (a) iso-butane, (b) trans-but-2-ene, (c) cis-but-2-ene, (d) but-1-ene
1
954
CHEM. COMMUN., 2003, 1954–1955
This journal is © The Royal Society of Chemistry 2003

4 3 2
Table 1 Ratios of the concentration of C products during the activation of MoO under an n-butane/H feed at 623 K
Thermodynamic ratio
at 623 K
1h
2h
6h
Steady-statea
Dehydrogenation
n-butane/but-1-ene
iso-butane/iso-butene
16000
0.045
8400
7000
17
6000 ± 800
500 ± 80
6600
470
0.91
Double-bond isomerisation
trans-but-2-ene/but-1-ene
cis-but-2-ene/but-1-ene
Skeletal isomerisation
iso-butene/but-1-ene
2.51
1.63
2.72
1.77
2.60
1.75
2.7 ± 0.4
1.8 ± 0.3
2.89
2.06
0.27
1.6 10
1.63
5600
3.3
120
1.2 ± 0.2
9.3 ± 1
8.12
1.73
6
n-butane/iso-butane
a
The average value for the three data points after 20 h is reported.
Katrib et al. have shown by XPS studies that MoO
formed in such conditions, possesses a metallic character.
Other experiments (to be reported elsewhere) have shown that a
pure MoO phase leads to the same equilibrium concentration of
linear butenes, without any iso-butene being formed. Regarding
this last point, Wehrer et al. have also shown that a pure MoO
2
, which is
6
2
2
phase is not active for n-hexane and hex-1-ene skeletal
isomerisation.12 The time-delayed formation of iso-butene and
iso-butane shows that the isomerising function is developed
subsequently, most likely related to the increase in sample
Brønsted acidity, as Matsuda et al.11 pointed out by propan-2-ol
dehydration experiments. It is not yet clear from which phase
the acidity originates, e.g. the oxycarbide or an amorphous
oxidic phase. However, the acidity is unlikely to be associated
2
with a pure MoO phase.
The decrease in iso-butene yield at reaction times greater than
five hours can probably be explained by the fact that the metallic
character of the sample increases much faster than the acidity.
The rate of formation of iso-butene is always slower than that of
linear butenes, but, initially, it is probably of the same order of
magnitude. Subsequently, the dehydrogenation/hydrogenation
is so much faster that the iso-butene concentration is depleted at
the benefit of that of iso-butane. This conclusion is supported by
the fact that, at the steady-state, n-butane is in thermal
equilibrium with the linear butenes and iso-butane is in
equilibrium with iso-butene (see Table 1). The increased
metallic character could be associated with the increase of the
Scheme 1 Reaction equilibria and rate-determining step taking place during
n-butane reaction over reduced MoO under steady-state conditions.
3
mechanisms and highlight rate-determining steps. In the present
case, this may result in a step improvement of the activity of a
most promising catalyst for alkane isomerisation.
specific surface area of the MoO
2
and the oxycarbidic phases
could exhibit a metal-
Notes and references
(
being an interstitial alloy, the MoO C
x y
13
‡ Part of this work was realised at the LERCSI-LCMC, University of
Strasbourg. The author gratefully acknowledges the French Ministry of
Research and Technology for financial support. The McClay Trust (School
of Chemistry, Belfast) is also acknowledged for financial support.
like character ). Therefore, the data reported here strongly
suggest that the rate-determining step of n-butane isomerisation
to iso-butane over this catalyst is the skeletal isomerisation of
linear butenes to iso-butene. A schematic representation of the
various equilibria and the rate-determining step taking place
1
2
R. Burch, J. Chem. Soc., Faraday I, 1978, 74, 2982.
G. A. Tsigdinos and W. W. Swanson, Ind. Eng. Chem. Res., 1978, 17,
over the reduced MoO
3
at steady-state conditions is given below
(see Scheme 1). The addition (as a mechanical mixture or as a
2
08.
support) of selective butene isomerisation catalysts such as the
zeolite FER or CoAPO-1114 could lead to much increased
reaction rates. This possibility is currently under investiga-
tion.
It is interesting to note that the rate-determining step in the
5 7
case of the isomerisation of C –C alkanes was proposed to be
3 A. P. E. York, C. Pham-Huu, P. Delgallo, E. A. Blekkan and M. J.
Ledoux, Ind. Eng. Chem. Res., 1996, 35, 672.
4 P. Delgallo, C. Pham-Huu, A. P. E. York and M. J. Ledoux, Ind. Eng
Chem. Res., 1996, 35, 3302.
5
6
7
T. Matsuda, K. Watanabe, H. Sakagami and N. Takahashi, Appl. Catal.
A, 2003, 242, 267.
A. Katrib, V. Logie, N. Saurel, P. Wehrer, H. Leflaive and G. Maire,
Surf. Sci., 1997, 377–379, 754.
C. Bouchy, C. Pham-Huu, B. Heinrich, C. Chaumont and M. J. Ledoux,
Appl. Catal. A, 2001, 215, 175; C. Bouchy, C. Pham-Huu, B. Heinrich,
C. Chaumont and M. J. Ledoux, J. Catal., 2000, 190, 92.
the dehydrogenation/hydrogenation process, as suggested by
5
propan-2-ol conversion data. This discrepancy could be related
to the difference of reactivity of the corresponding alkenes, i.e.
the isomerisation of butene is less favourable, possibly because
of the higher energy requirements associated with C
4
carbonium
8 F. C. Meunier, J. P. Breen and J. R. H. Ross, Chem. Commun., 1999,
259.
intermediates.14 Also, inferences on the rates of isomerisation
and hydrogenation/dehydrogenation of an alkane from data on
propan-2-ol conversion may not be fully quantitative. This
underlines that the rate-determining step for the reaction of C5+
chains may be different and needs to be further investigated.
In conclusion, this paper shows that the quantitative analysis
of trace amounts of reaction products (the butene concentrations
are in the order of hundreds of ppm) and the related
thermodynamics is a powerful tool to elucidate reaction
9
J. P. Breen, F. C. Meunier and J. R. H. Ross, Chem. Commun., 1999,
247.
0 A. Katrib, P. Leflaive, L. Hilaire and G. Maire, Catal. Lett., 1996, 38,
5.
1 T. Matsuda, Y. Hirata, H. Sakagami and N. Takahashi, Chem. Lett.,
997, 1261.
2 P. Wehrer, C. Bigey and L. Hilaire, Appl. Catal. A, 2003, 243, 109.
2
1
1
1
9
1
13 S. T. Oyama, Catal. Today, 1992, 15, 179.
14 J. Houzvicka and V. Ponec, Catal. Rev., 1997, 39, 319.
CHEM. COMMUN., 2003, 1954–1955
1955