1-Hexene Oligomerization in Liquid, Vapor, and Supercritical Phases over Beidellite and Ultrastable Y Zeolite Catalysts

Article abstract of DOI:10.1006/jcat.1998.2250

1-Hexene was oligomerized at 200°C and a pressure of 5 MPa in a down-flow fixed-bed tubular reactor filled with beidellite or ultrastable Y zeolite catalysts. Vapor, liquid, and supercritical states of the reacting hydrocarbons in the reactor tube were established by using propane, pentane, octane, and dodecane as solvents. The initial activity, stability with operation time, and selectivity are very dependent on the physical state of the hydrocarbons. Highest activity and stability are reached in the liquid phase using octane and dodecane solvent. The chain length of the solvent has a strong influence on the deactivation of the catalyst, on the oligomerization selectivity and on the formation of C₆ saturates and cracked products.

Full text of DOI:10.1006/jcat.1998.2250

JOURNAL OF CATALYSIS 179, 477–482 (1998)
ARTICLE NO. CA982250
1-Hexene Oligomerization in Liquid, Vapor, and Supercritical Phases
over Beidellite and Ultrastable Y Zeolite Catalysts
Je´roˆme P. G. Pater, Pierre A. Jacobs, and Johan A. Martens1
Departement Interfasechemie, Centrum voor Oppervlaktechemie en Katalyse, K.U. Leuven, Kardinaal Mercierlaan 92, B-3001, Hevelee, Belgium
Received March 13, 1998; revised July 29, 1998; accepted July 31, 1998
Under such reaction conditions, the alkene undergoes
1-Hexene was oligomerized at 200ꢀC and a pressure of 5 MPa
in a down-flow fixed-bed tubular reactor filled with beidellite or ul-
trastable Y zeolite catalysts. Vapor, liquid, and supercritical states
of the reacting hydrocarbons in the reactor tube were established
by using propane, pentane, octane, and dodecane as solvents. The
initial activity, stability with operation time, and selectivity are very
dependent on the physical state of the hydrocarbons. Highest ac-
tivity and stability are reached in the liquid phase using octane and
dodecane solvent. The chain length of the solvent has a strong in-
fluence on the deactivation of the catalyst, on the oligomerization
selectivity and on the formation of C6 saturates and cracked prod-
c
oligomerization, isomerization, cracking to a mixture of in-
termediate olefins that is independent of the nature of the
feedstock, and copolymerization to give a product contain-
ing all carbon numbers (13). Tabak claimed the used of
H-ZSM-12 zeolite catalysts in a liquid phase process at low
temperatures. In these processes, 1-hexene and 1-octene
underwent selective dimerization reactions (14).
Kerosene consumption and, especially in Europe, diesel
consumption are increasing. One possibility to produce
these fractions is through oligomerization of C=5 -C7= FCC
alkenes via acid catalysis. In this paper, we have studied the
oligomerization of 1-hexene as a model compound repre-
sentative of this fraction over two solid acid catalysts, acid
beidellite, and ultrastable Y zeolite. We compared gas, liq-
uid, and supercritical reaction conditions in a search for
catalyst operation conditions giving high conversion, slow
deactivation, and high oligomerization selectivity.
ucts.
ꢁ 1998 Academic Press
Key Words: 1-Hexene oligomerization; hydrogen transfer reac-
tions; ultrastable Y zeolite; beidellite; solvent effect.
1. INTRODUCTION
Acid catalyzed oligomerization of C=3 -C4= alkenes pro-
duced by FCC (fluid catalytic cr=acking) into high-quality
gasoline-range iso-olefins (C=6 -C10) are established petro-
chemical processes (1–3). The oldest process using silica
supported phosphoric acid dates back to 1935 (1). Medium-
pore, shape-selective acid zeolites such as H-ZSM-5 have
superior activities and durabilities in the processes desig-
nated as “MOGD” (Mobil olefins to gasoline and distil-
late) (4–8). Large-pore zeolites with high silica-to-alumina
ratio (9–11) and cation-exchanged clays (12) can be use=d
to oligomerize terminal and internal long-chain C=10-C20
n-alkenes into synthetic lubricant base stocks. In contrast
to the short and long n-alkenes, literature data on the
oligomerization of alkenes with carbon numbers in the
range C=5 -C=7 is more scarce. In patents dealing with alkene
oligomerization, processes and catalysts are often claimed
to be suitable for C2-C12 feedstocks, but in these patents, ex-
amples on C=5 -C=7 alkenes are seldomly provided. Garwood
reported the conversion of C2-C10 alkenes on H-ZSM-5 ze-
olite in a vapor phase process at temperature of 200–260ꢀC.
2. EXPERIMENTAL
Beidellite was synthesized according to a procedure re-
ported by Le Dred and Baron (15). The as-synthesized bei-
dellite sample was NH4-exchanged and freeze-dried to ob-
tain a powder with B.E.T. specific surface area of 62 m2 ꢂ gꢃ1
.
The sample does not contain any microporosity, has a meso-
pore volume of0.159cm3 ꢂ gꢃ1 asderived from N2 physisorp-
tion isotherms at 77 K, and a Si/Al atomic ratio in the tetra-
hedral layer, assuming absence of Al-O-Al linkages, was 4.7
(15). The ultrastable Y zeolite powder sample with Si/Al
atomic ratio of ca 15 was from PQ (code name CBV-720).
Compared to beidellite, the US-Y zeolite presents a higher
B.E.T. specific surface area (658 m2 ꢂ gꢃ1), with micropore
and mesopore volumesof0.253cm3 ꢂ gꢃ1 and 0.122cm3 ꢂ gꢃ1
,
respectively. This ultrastable Y zeolite and pillared versions
of beidellite present a similar acidity and catalytic activity
in hydroisomerization–hydrocracking of decane (16–17).
Catalyst powders were compressed into wafers that were
crushed and sieved to obtain particles with diameters of
250–500 ꢀm for use in the catalytic experiments. The
1 Corresponding author. E-mail: Johan.Martens@agr.kuleuven.ac.be.
477
0021-9517/98 $25.00
c
Copyright ꢁ 1998 by Academic Press
All rights of reproduction in any form reserved.

478
PATER, JACOBS, AND MARTENS
TABLE 1
Reaction Conditions
Exp.
run
Flow-rate
(golefin ꢂ gꢃca1ta ꢂ hꢃ1
Solvent
(wt% in feed)
T
(ꢀC)
P
Catalyst
)
(MPa)
Physical state
A
B
C
D
Beidellite
Beidellite
Beidellite
Beidellite
3
3
3
3
Propane (70% )
Pentane (70% )
Octane (70% )
Dodecane (70% )
200
200
200
200
5
5
5
5
Vapor
Supercritical
Liquid
Liquid
E
F
G
H
US-Y
US-Y
US-Y
US-Y
1
1
3
3
Pentane (90% )
Pentane (90% )
Octane (70% )
Dodecane (70% )
200
200
200
200
1
5
5
5
Vapor
Supercritical
Liquid
Liquid
catalyst pellets were activated in situ in the reactor tube by
a calcination under flowing oxygen (3 cm3 ꢂ sꢃ1 ꢂ gꢃ1) during
5 h at 450ꢀC.
3. RESULTS AND DISCUSSION
3.1. Beidellite Catalyst
1-Hexene (97% ), hexane (>99% ), heptane (>99% ), oc-
tane (>99% ), and dodecane (>99% ) were from Acros,
propane (>99.9% ) from Alphagaz, and pentane (>99% )
from Lab-Scan Analytical Science.
The oligomerization of 1-hexene was performed on the
beidellite catalyst at a temperature of 200ꢀC, a pressure of
5 MPa, and either propane, pentane, octane, or dodecane
as solvent, representing 70 wt% of the mixture (Table 1).
The physical state of the hydrocarbon mixtures under reac-
tion conditions can be estimated from the phase diagrams
of the n-alkane solvents and considering a partial pressure
of 3.5 MPa (Fig. 1). With propane, the vapor phase is estab-
lished in the reactor; with pentane, the supercritical state.
With octane and dodecane, the reactor is operating under
liquid phase conditions. The reaction products obtained
were divided into four groups: 1-hexene isomers, oligo-
mers, cracked products, and C6-hydrogenated products. The
following isomers of 1-hexene were formed: t-2-hexene,
c-2-hexene, 2-methyl-2-pentene, 3-methyl-t-2-pentene, and
3-methyl-c-2-pentene. In view of the purpose of our study,
all C=6 products were lumped and considered as uncon-
verted “hexene.” The C6 hydrogenated products formed
were hexane, 2-methylpentane, and 3-methylpentane. The
oligomers formed were dimers (C=12) essentially, together
with trace quantities of trimers (C=18). Owing to the large
quantities of C12 formed, it was not possible to separate and
analyze the small amount of C=9 -C1=1 products accurately.
Therefore, the C=9+ fraction has been lumped and con-
sidered as oligomers, realizing that C=9 -C1=1 belong strictly
speaking to the cracked product fraction. The remaining
products have carbon numbers of C3-C5, C7, and C8 and are
regarded as cracked products. The C4 and especially the C5
fraction contained significant amounts of alkanes next to
alkenes.
The 1-hexene conversions were carried out in a down-
flow fixed-bed tubular microreactor. The stainless steel tube
holding 0.5–3 g of catalyst between two plugs of quartz wool
had an internal diameter of 1 cm and a length of 10 cm and
was mounted in a tubular furnace. The dead volume inside
the reactor tube was filled with glass beads with a same
diameter as the catalyst particles. The temperature in the
catalyst bed was monitored with an axially mounted ther-
mocouple. The feedstock mixtureswere prepared bymixing
pure compounds in a pressure vessel equipped with stirrer.
Heptane was used as internal standard. The feed was deliv-
ered to the reactor through a mass flow controller for liquids
(Rosemount/Brooks Instruments B.V., type 5881), and the
pressure in the reactor regulated with a back-pressure reg-
ulator. The reactor was pressurized with nitrogen before
feeding the hydrocarbons.
The reaction products were sampled at high pressure
using a sampling valve with an internal volume of 0.1 ꢀl,
located upstream of the back-pressure regulator, and trans-
ferred by means of a heated transfer line to a gaschro-
matograph (HP 5890 Series II) equipped with a fused silica
WCOT column (CP-SiL 5 CB from Chrompack; length of
50m, internaldiameter of0.32mm;film thicknessof5,0ꢀm)
and F.I.D. detector.
Liquid product fractions were collected in a gas–liquid
separator at 15ꢀC and atmospheric pressure. The skele-
tal branching of the oligomers was analyzed using a cata-
lytic hydrogenation module (modified TCT injection sys-
tem from Chrompack using Pd catalyst at 180ꢀC), mounted
at the inlet side of the same type of gaschromatograph as for
the on-line analysis. Identification of the products was done
using reference samples obtained by hydroisomerization of
dodecane and using GC/MS.
The evolution with time-on-stream of the hexene con-
version over the beidellite catalyst is shown in Fig. 2. The
initial hexene conversion decreases in the solvent order:
dodecane (89% ) ꢄ octane (86% ) > pentane (74% )
> propane (30% ).
[1]

1-HEXENE OLIGOMERIZATION OVER BEIDELLITE AND USY
479
FIG. 1. Physical state of alkanes depending on pressure and temperature.
In the vapor phase, a low and stable conversion of ca 30% catalyst were considerably better. In about 16 h on-stream,
is obtained in the first 16 h. Possibly, the catalyst underwent the hexene conversion decreased from ca 90% to 60% in
a fast deactivation in the first hour of the experiment. The the presence of octane, and from ca 90% to 70% in pres-
black color of the catalyst after the test was indicative of ence of dodecane, indicating that the stability of the catalyst
severe coking. In the supercritical phase and in presence of is improved when using a heavier alkane solvent. Catalysts
pentane, the catalyst was initially more active but a rapid operated under the liquid phase conditions had a yellow
deactivation took place in the first 16 h. In the experiments color after the testing, suggesting less coke deposition.
in the liquid phase, in presence of octane or dodecane sol-
The selectivities obtained under the different operation
vents, the initial conversion as well as the stability of the conditions of the beidellite catalyst at three conversion
FIG. 2. Evolution with time-on-stream of the hexene conversion in propane, pentane, octane, and dodecane over beidellite catalyst: ᭜, dodecane;
᭿, octane; ଙ, pentane; ᭹, propane.

480
PATER, JACOBS, AND MARTENS
TABLE 2
Selectivity of Hexene Conversion and Dimer Skeletal Distribution Obtained on Beidellite Catalyst
Run
A
B
C
D
Solvent
Physical state
Propane
Vapor
Pentane
Supercritical
Octane
Liquid
Dodecane
Liquid
Time-on-stream (h)
Hexene conversion (wt% )
1.5
30
15
34
1.5
74
15
33
1.5
86
9
74
1.5
86
13.5
76
Cracking
C6 saturates formation
Oligomerization
6
18
76
8
26
66
Cracking
C6 saturates formation
Oligomerization
4
34
62
3
19
78
6
22
72
Cracking
C6 saturates formation
Oligomerization
3
75
22
3
76
21
1
71
28
Skeletal dimer
distribution (wt% )
Monobranched + Normal
Dibranched
7
83
10
7
89
4
18
79
3
24
70
6
Tribranched
Monobranched dimer
distribution (wt% )
5-Ethyldecane
8.6
6.7
5-Methylundecane
4-Ethyldecane
2 & 4 Methylundecane
+ 3 Ethyldecane
3-Methylundecane
n-Dodecane
58.0
17.3
7.8
59.9
18.9
6.4
6.7
1.7
6.2
1.8
levels reached after specific times-on-stream in the experi-
mentsofFig. 1(ca 85% , 75% , and 30% ) are given in Table 2.
In the liquid phase with octane and dodecane solvent, the
selectivities did not change substantially with the time span
of the experiment and upon deactivation (Table 2). Com-
paring the solvents at a same conversion level obtained
after different degrees of deactivation, the selectivity for
oligomerization decreases in the solvent order (Table 2):
octane > dodecane > pentane > propane.
[2]
The selectivity for C6 saturates formation increases in
the opposite order, while the cracking is little influenced
(Table 2).
The C12 fraction is mainly composed of dibranched
isomers, together with monobranched isomers and small
amounts of tribranched isomers (Table 2). The mono-
branched C12 fraction, collected in the experiments with
propane and octane, was analyzed in detail using the GC
with sample hydrogenation facility (Table 2). The main
products in this fraction are 5-methylundecane and 4-ethyl-
decane. These skeletal structures are primary products re-
sulting from the addition of 1-hexene to hex-2-yl and hex-
3-yl cations, respectively (Scheme 1). Isomers with other
SCHEME 1

1-HEXENE OLIGOMERIZATION OVER BEIDELLITE AND USY
481
FIG. 3. Evolution with time-on-stream of the hexene conversion in octane and dodecane over ultrastable Y zeolite: ᭿, octane; ᭜, dodecane.
branching positions are formed in small quantities indicat- sure of 1 MPa leads to vapor phase conditions (Fig. 1) and a
ing that alkyl shifts are relatively slow under these con- low initial hexene conversion (45% , Table 3). As with bei-
ditions. The small quantity of n-dodecane formed by un- dellite, the activity of the ultrastable Y zeolite decreases in
branching of monobranched isomers is another indication the following order of physical states for the hydrocarbons:
that secondary isomerization is insignificant. The occur-
rence of secondary isomerization does not seem to be much
dependent on the reaction conditions (Table 2). The com-
position of the monobranched C12 products obtained in
propane solvent in the vapor phase and at 30 to 34% con-
version is very similar to that obtained in octane solvent in
the liquid phase and at 86 to 74% conversion.
In the dibranched fraction, two main products could be
identified, viz. 5,6-dimethyldecane and 4-ethyl,5-methyl-
nonane. These are expected primary dimerization products
(Scheme 1). Quantification of the dibranched C12 compo-
sition was, however, not possible given the severe overlap-
ping of GC peaks.
liquid phase > supercritical phase > vapor phase. [3]
Competitive physisorption phenomena between reagent,
product, and solvent molecules could be at the origin of
this influence of the physical state. Propane and pentane
molecules with a lower molecular weight than the hexene
dimers are less effective for evacuation of the oligomeriza-
tion products from the catalyst and cause fast deactivation.
On ultrastable Y, the highest oligomerization selectivity
was obtained in octane (62% at 93% conversion, Table 3).
In dodecane and at a similar conversion, it was 52% only
(Table 3). This difference is due to a higher selectivity for C6
saturates (21% in dodecane, compared to 13% in octane,
Table 3), while the cracking selectivity is little influenced.
It can be concluded that the reaction conditions have a
strong impact on saturates formation and deactivation, but
much lesson the oligomerization mechanism and secondary
isomerization of the hexene dimers.
TABLE 3
3.2. Ultrastable Y Zeolite Catalysts
Selectivity of Hexene Conversion on Ultrastable Y Zeolite
The ultrastable Y zeolite catalyst was used in a variety
of reaction conditions in order to verify whether the obser-
vations made with beidellite can be generalized (Table 1).
In the experiments in octane and dodecane at 200ꢀC and a
pressure of 5 MPa, high hexene conversions exceeding 90%
were reached initially (Fig. 3 and Table 3). In supercritical
pentane at the same pressure and temperature, the hexene
conversion reaches initially 75% only, even at three times
lower space velocity (Table 3). Using pentane and a pres-
Run
Solvent
Physical state
E
F
G
H
Pentane
Vapor Supercritical Liquid
Pentane
Octane Dodecane
Liquid
Time-on-stream (h)
Hexene conversion (wt% )
1.5
45
1.5
75
1.5
93
1.5
96
Cracking
C6 saturates formation
Oligomerization
12
62
26
14
40
46
23
13
62
27
21
52

482
PATER, JACOBS, AND MARTENS
There is one main difference between the behavior of
ACKNOWLEDGMENTS
the zeolite and the clay. The ultrastable Y zeolite deacti-
vates faster in dodecane than in octane (Fig. 3), while the
opposite is observed for beidellite (Fig. 2). The zeolite con-
tains micropores and mesopores, while the clay material is
essentially meso and macroporous. On beidellite, as well
as ultrastable Y, hydrogen transfer reactions to produce
saturates are more important in dodecane than in octane
(Tables 2 and 3). It can be speculated that in the zeolite,
the hydrogen deficient molecules cause deactivation by mi-
cropore blockage, whereas in the clay with its larger pores
they are more easily desorbed. It remains to be explained
why the hydrogen transfer reactions are less important in
octane, compared to dodecane, and this is independent of
the nature of the catalyst.
This work has been sponsored by Total Raffinage Distribution S.A.
Stimulating discussions with J. M. Colin, R. Loutaty, A. Pantazidis,
G. Szabo (CERT TOTAL, Harfleur-France), and C. Naccache (IRC,
Villeurbanne-France) were greatly appreciated.
REFERENCES
1. Ipatieff, V. N., Corson, B. B., and Egloff, G., Ind. Eng. Chem. 27, 1077
(1935).
2. Jones, E. K., in “Advances in Catalysis” (Frankenburg & Komarewsky,
Eds.), Vol. VIII, p. 219. Academic Press, New York, 1956.
3. Oblad, A. G., Mills, G. A., and Heinemann, H., in “Catalysis” (Emmet,
Ed.), Vol. VI, p. 341. Reinhold, New York, 1958.
4. van den Berg, J. P., Wolthuizen, J. P., and van Hooff, J. H. C., J. Catal.
80, 139 (1983).
5. van den Berg, J. P., Wolthuizen, J. P., Clague, A. D. H., Hays, G. R.,
Huis, R., and van Hooff, J. H. C., J. Catal. 80, 130 (1983).
6. Quann, R. J., Green, L. A., Tabak, S. A., and Krambeck, F. J.,
Ind. Eng. Chem. Res. 27, 565 (1988).
4. CONCLUSIONS
7. Gricus Kofke, T. J., and Gorte, R. J., J. Catal. 115, 223 (1989).
8. Sealy, S. J., Fraser, D. M., Mo¨ller, K. P., and O’Connor, C. T., Chem.
Eng. Sci. 49, 3307 (1994).
The activity, stability, and selectivity of beidellite and ul-
trastable Y zeolite in 1-hexene oligomerization are strongly
influenced by the physical state of the hydrocarbons under
9. Knifton, J. F., and Sanderson, J. R., Catal. Lett. 28, 223 (1994).
reaction conditions. With both catalysts, liquid phase re- 10. Dai, P-S. E., Sanderson, J. R., and Knifton, J. F., in “Studies in Surface
Sciences and Catalysis” (J. Weitkamp et al., Eds.), Vol. 84, p. 1701.
Elsevier, Amsterdam, 1994.
11. Stepanov, A. G., Luzgin, M. V., Romannikov, V. N., and Zamaraev,
action conditions achieved by using octane and dodecane
solvents are most appropriate with respect to catalytic ac-
tivity, oligomerization selectivity, and stability. Under liq-
K. I., Catal. Lett. 24, 271 (1994).
uid phase conditions, the molecular weight of the solvent
has a strong influence on the selectivity and the stability of
the catalyst. On beidellite as well as on ultrastable zeolite
Y, hydrogen transfer reactions resulting in the formation
of C6 saturates are the main side reactions. These hydro-
gen transfer reactions are more abundant in dodecane than
in octane. Deactivation seems to be linked to the forma-
tion of large hydrogen deficient molecules, blocking micro-
and mesopores in ultrastable Y zeolites but not the larger
pores in beidellite. It explains why the beidellite catalyst is
most stable in dodecane, the ultrastable Y zeolite catalyst in
octane.
12. Muthukumaru Pillai, S., and Ravindranathan, M., J. Chem. Soc. Chem.
Commun., 1813 (1994).
13. Garwood, W. E., in “Intrazeolite Chemistry” (D. Stucky and F. G.
Dwyer, Eds.), A.C.S Symp. Ser., Vol. 218, p. 383. Am. Chem. Soc.,
Washington, DC, 1983.
14. Tabak, S. A., U.S. Patent 4,254,295 (1981).
15. Miehe, J., Le Dred, R., Baron, J., and Saehr, D., in “Synthesis of Porous
Materials” (M. L. Occelli and H. Kessler, Eds.), p. 523. Dekker, New
York, 1996.
16. Poncelet, G., and Schutz, A., in “Chemical Reactions in Organic
and Inorganic Constrained Systems” (R. Setton, Ed.), p. 165. Reidel,
Dordrecht, 1986.
17. Molina, R., Moreno, S., Vieira-Coelho, A., Martens, J. A., Jacobs,
P. A., and Poncelet, G., J. Catal. 148, 304 (1994).