Selective Isomerization of n-Butene into Isobutene over Deactivated H-Ferrierite Catalyst: Further Investigations

Full text of DOI:10.1006/jcat.1997.1656

JOURNAL OF CATALYSIS 169, 397–399 (1997)
ARTICLE NO. CA971656
NOTE
Selective Isomerization of n-Butene into Isobutene over Deactivated
H–Ferrierite Catalyst: Further Investigations
During the past decade, there was a renewed interest and that most of the porous volume of the solid is blocked
in the skeletal isomerization of n-butene due to the de- by carbonaceous desposits; in addition, they observed that
mand for isobutene used to produce methyltertiarybutyl after coking the remaining number of acid sites, indirectly
ether (MTBE).
probed by ammonia or 1-butene TPD is quite low. Another
Various acid-type catalysts have been used such as group (10) has proposed that on the aged ferrierite catalyst,
oxides like MoO3 or WO3 or halogenated alumina and ze- the active site for the squeletal isomerization of butene is
olites.
a carbenium ion trapped within the carbonaceous residues
Recently, Cheng and Ponec (1) addressed the question of formed in the ferrierite pores and responsible for its deac-
whether or not the selective isomerization of n-butene into tivation.
isobutene over these catalysts proceeds through a common
intermediate.
If the transformation of n-butene into isobutene via a
monomolecular reaction mechanism is catalyzed by a pro-
In a recent publication, we have shown that on MFI-type ton, it is generally accepted that this transformation in-
catalysts, two different pathways operate for the isomeriza- volves an unstable primary carbenium ion intermediate via
tion ofn-butene into isobutene, one via a bimolecular mech- a protonated cyclopropane intermediate:
anism and another one via a monomolecular mechanism,
the last one being favored by high reaction temperature
and low pressure (2). Identical conclusions were reached
for nondeactivated ferrierite (2).
As has been clearly shown in the literature, medium pore
zeolites like ZSM-22 (3) or ZSM-23 (4) are selective cata-
lysts for this reaction.
More recently, a large number of papers were devoted
to the study of ferrierite catalysts (5) which have been
patented as effective catalysts for this isomerization reac-
tion (6).
For ferrierite materials, it is observed that the isobutene
selectivity increases with time on stream (5).
C
C

H+
+
–+ – –
–
– –+
%H e
== – –
C
C C C ꢀ C C C C ꢀ C
C C → C C C
[1]
This type of cyclopropanic intermediate has been first pro-
posed by Brouwer and Hogeveen (11).
By contrast, for a larger olefin like pentene, the mono-
molecular reaction occurs via a secondary carbenium ion.
H+
== – – –
–+ – – –
C
C C C C ꢀ C C C C C
C
C

+
–
– –+ –
%H e
For these narrow pore materials, it has been suggested
that a bimolecular mechanism is operating for fresh cata-
lysts (7) acidic OH groups being the active sites, but for aged
materials, which are highly selective, the reaction probably
occurs via a monomolecular mechanism (8).
–
ꢀ C C
C C → C C C C
[2]
The reaction rate via [1] should be much smaller than via
[2] since in the first scheme the reaction is occurring via
secondary–primary carbenium ions, while for the second it
is via secondary–secondary carbenium ions.
If the reaction is catalyzed by a carbonaceous carbenium
ion (10)
In a recent note, Me´riaudeau et al. (9), using 13C-
labeled butene have demonstrated that for a selective
ferrierite (aged solid), the monomolecular mechanism is
operating.
Despite the fact that this reaction, over ferrierite cata-
lyst, has been extensively studied several questions remain
unsolved, among them, that of the nature of the active site
and its location. Concerning the nature of the active site,
Xu et al. (5) have shown that the aging of the solid with time
on stream leads to a large increase in isobutene selectivity
R
 2
+
–
R1
C

R3
the formation of isobutene would occur via secondary–
secondary carbenium ions as proposed in (10).
397
0021-9517/97 $25.00
c
Copyright ꢀ 1997 by Academic Press
All rights of reproduction in any form reserved.

398
NOTE
TABLE 1
R
R
R
 2
2
2
+
– == –
–
– – –+ –
– – –+ –
Catalytic Properties of Fresh and Aged Solid for 1-C4H8 and
C C C C + R1 C → R C C C C → R C C C C





1
1
2-(cis + trans)-C5H10 Transformation at 673 K
R3
R3 C
R3
C
R
Fresh catalyst Aged catalysta
 2
+
–
– ==
→ R1 C + C C
C
[3]


c
n-Butene conversion,b (WHSV, hꢁ1
)
4% (3,400)
3% (4,700)
2% (45)
R3
C
Isobutene selectivityd
45%
92%
Isobutene rate formatione
1.1 (4% )
1.1 (3% )
0.015 (2% )
If the site responsible for the skeletal isomerization of pen-
tene is (over ferrierite covered by coke) again the previ-
ously cited carbenium ion (R1) (R2) (R3) C+, the pathway
proposed for n-butene would also prevail and again would
involve a secondary carbenium ion intermediate.
Thus if the skeletal isomerization of C4 and C5 olefins
is catalyzed by a carbonaceous carbenium ion, the isomer-
ization rates of n-butene and n-pentene are expected to
be similar, while by contrast for H+ catalyzed monomolec-
ular reaction, the rate of n-pentene isomerization should
be higher than that of n-butene. In order to discriminate
between H+ or (R1) (R2) (R3) C+ as the active site op-
erating in coked ferrierite in the skeletal isomerization of
linear olefins (butene, pentene), we have measured the re-
spective rates of 1-butene isomerization and of 2-cis- and
trans-pentene isomerization over coked ferrierite.
(butene conversion)b
n-Pentene conversion,b (WHSV, hꢁ1
)
12% (30,000)
8% (48,000)
3% (1700)
c
Isopentene selectivityd
Isopentene rate formatione
(pentene conversion)b
92%
98%
45 (12% )
48 (8% )
0.7 (3% )
Note. Polefin = 26 kPa; complement to atmospheric pressure, N2.
a Solid aged for 16 h (see Fig. 1).
b
For butenes (or pentenes) the conversion is defined as (Cin ꢁ Cout)/
Cin ꢂ 100, with Cin being the number of moles of 1-C4H8, Cout = 6(1-
C4H8 + 2-cis C4H8 + 2-trans-C4H8). Same assumptionsfor linear pentenes.
Contact times chosen to have conversion ꢃ 12% .
c WHSV adjusted to have conversions ꢃ 12% .
d
Carbon basis, all linear butene (1-butene, 2-cis- and 2-trans-butene)
are considered as reactants. Same assumption for pentene.
e Rate in mol ꢄ hꢁ1 gꢁ1 catalyst.
For this study, we have used a ferrierite, synthesized with-
out templatingagent, havingSi/Al = 10. It hasbeen checked
by XRD that the ferrierite phase is pure and highly crys-
talline. After having exchanged the Na form with NH4Cl,
the protonic form was obtained by calcining the solid un-
der N2 at 773 K. 1-Butene reaction at 673 K, PC H = 26 kPa,
Part of this catalyst was pressed in a form of a disk for
IR studies and another part was mechanically mixed with
silica (Aerosil, Degussa) and loaded in a microreactor for
catalytic studies.
Additional experiments have shown that the contact with
air of the catalyst at RT does not modify the catalytic prop-
erties of the aged solid.
The catalyst which has been aged during the reaction
of 1-butene and was in a selective state was then used
without further treatment for a series of 1-butene and
2-(cis + trans)-pentene isomerization reaction experiments.
The reactant space velocity was selected such that the
reaction kinetics were not controlled by external mass
transfer.
4
8
WHSV = 6 hꢁ1 was run for 16 h until reaching an isobutene
selectivity of 90% for a conversion of 40% (see Fig. 1); the
reaction of 1-butene was then interrupted and the catalyst
cooled to RT and the reactor was open to air.
In Table 1 are listed the results of the isomerization ex-
periments performed on the fresh ferrierite and the aged
ferrierite, coked as indicated above.
The conversions obtained are low and of same magnitude
for the aged solid reacted with either butene or pentene. By
contrast, for the fresh solid used for n-pentene isomeriza-
tion it has not been possible to have a very low conversion
even for this highest WHSV but since the isopentene rate
formation is nearly constant when calculated with conver-
sion of 12 and 8% , this strongly suggest that the rate so
calculated is very close to the true rate value.
From Table 1, two features are clearly apparent:
—the rate of isobutene (or isopentene) formation is
strongly depressed on the aged ferrierite;
FIG. 1. Transformation of 1-butene on ferrierite. T = 673 K; WHSV =
6 hꢁ1; PC 8 = 26 kPa; complement to atmospheric pressure, N2.
H
4

NOTE
399
—the rate of isobutene formation (starting from 1-
ACKNOWLEDGMENT
butene) is much smaller than that of isopentene (starting
from 2-pentene) on both fresh and aged ferrierites.
Total Raffinage Distribution is acknowledged for funding.
REFERENCES
It is improbable that the mechanism involving the car-
bonaceous carbenium ion active site for the skeletal iso-
merization of n-butene over aged ferrierite (10) would
not operate for n-pentene isomerization. Since our results
have clearly shown that on aged ferrierite n-pentene is iso-
merized much faster than n-butene, it appears also im-
probable that the isomerization steps involve the same
skeletal intermediates namely secondary–secondary carbe-
nium ions. This precludes the carbonaceous carbenium ion
residues, on the aged ferrierite, from being the active site for
the n-butene isomerization. The selective isomerization of
n-butene is probably catalyzed by proton acid sites on the
internal surface of ferrierite crystals, with a monomolecular
reaction path via cyclopropane intermediates prevailing on
aged catalysts (9).
1. Cheng,, Z. X., and Ponec,, V., J. Catal. 148, 607 (1994).
2. Me´riaudeau, P., Vu, T. A., Le, N. H., and Naccache, C., “Studies
in Surface Science and Catalysis” (H. Chon et al., Eds.). Vol. 105C,
p. 1373. Elsevier, Amsterdam, 1966.
3. Simon, M. W., Suib, S. L., and O’Young, C. L., J. Catal. 150, 34 (1994).
4. Xu, W. Q., Yin, Y. G., Suib, S. L., and O’Young, C. L., J. Catal. 34
(1994).
5. Xu, W. Q., Yin, Y. G., Suib, S. L., Edwards, J. C., and O’Young, C. L.,
J. Phys. Chem. 99, 9433 (1995).
6. Lyondell, U.S. Patent 711,041, 1991.
7. Guisnet, M., Andy, P., Gnep, N. S., Benazzi, E., and Travers, C., J. Catal.
158, 551 (1996).
8. O’Young , C. L., Pellet, R. J., Casey, D. G., Ugolini, J. R., and Sawicki,
R. A., J. Catal. 151, 467 (1995).
9. Me´riaudeau, P., Bacaud, R., Le, N. H., and Vu, A. T., J. Mol. Catal.
A Chem. 110, L177 (1996).
10. Guisnet, M., Andy, P., Gnep, N. S., Travers, C., and Benazzi, E.,
J. Chem. Soc., Chem. Commun. 1685 (1995).
The decrease in the butene conversion, as a function of
time on stream, is due to two phenomena induced by coke
deposits:
11. Brouwer, D. M., and Hogeveen, H., Prog. Phys. Org. Chem. 9, 179
(1972).
—a partial pore blocking as evidenced by Xu et al. (5);
—a partial site poisoning as indirectly evidenced by
butene TPD (5). Direct evidence of this was obtained by
using IR spectroscopy: we have observed that the vibra-
tion at 3600 cmꢁ1 attributed to acidic OH groups is reduced
by a factor of 2.4 after 16 h on stream. Indeed, the rate of
isobutene (or isopentene) formation is reduced by a much
larger factor (73 for isobutene and 64 for isopentene) sug-
gesting that either part of the acidic OH groups are no more
accessible to reactants and/or that the rate of the isomer-
ization reaction is diffusion controlled.
P. Me´ riaudeau
V. A. Tuan
N. H. Le
G. Szabo1
Institut de Recherches sur la Catalyse
CNRS, 2 avenue Albert Einstein
69626 Villeurbanne, France
Received November 1, 1996; revised February 10, 1997; accepted
February 13, 1997
Further work is needed to clarify this question.
1 Total Raffinage Distribution BP27 76700 Harfleur, France.