Decalin and tetralin as probe molecules for cracking and hydrotreating the light cycle oil

Article abstract of DOI:10.1006/jcat.2001.3181

VGO and residual oil are cracked in FCC units to liquid fuels (gasoline and LCO) for automotive fuels and short-chain olefins that are used as feedstocks in petrochemistry. The cracking of tetralin and decalin was performed over a set of zeolites with medium-sized (ZSM-5, MCM-22, ITQ-2), large (USY, Beta), ultralarge pores (UTD-1), and mesoporous MCM-41 at 723 K to establish the effect of the pore topology of the catalyst on product distribution. These molecules were selected as probe molecules because they indicate which catalyst is the best for cracking or hydrotreating the LCO fraction, which is obtained directly from FCC units. Pore size and topology had a strong effect on diffusion and on activity and selectivity in reactions such as ring opening, dealkylation, transalkylation, hydride transfer, and coke formation. The products in decalin cracking at 723 K include methane, ethylene, propane, propylene, tetralin, naphthalene, etc. A good LCO cracking catalyst for the production of naphthenes should involve a large-pore zeolite for primary cracking. Beta was a very suitable zeolite for this, together with ZSM-5 that can crack the smaller products formed after the primary cracking on Beta zeolite to give high yields of propylene and ethylene.

Full text of DOI:10.1006/jcat.2001.3181

Journal of Catalysis 200, 34–44 (2001)
doi:10.1006/jcat.2001.3181, available online at http://www.idealibrary.com on
Decalin and Tetralin as Probe Molecules for Cracking
and Hydrotreating the Light Cycle Oil
A. Corma,�,1
V. Gonz a´ lez-Alfaro,† and A. V. Orchill e´ s†
�
Instituto de Tecnolog ´ı a Qu ´ı mica, UPV-CSIC, Universidad Polit e´ cnica de Valencia, Avda. de los Naranjos s/n, 46022 Valencia, Spain; and
†
Departamento de Ingenier ´ı a Qu ´ı mica, Universitat de Val e` ncia, Dr. Moliner 50, 46100 Burjassot, Valencia, Spain
Received July 11, 2000; revised October 23, 2000; accepted January 15, 2001; published online April 11, 2001
cetane number) cause rough engine operation, misfiring,
Cracking of tetralin and decalin was carried out over several incomplete combustion, and poor startability. Cetane (n-
zeolites to establish the effect of the pore topology of the cata- hexadecane) has a short delay period during ignition and is
lyst on product distribution. These molecules were chosen as probe
molecules, because they indicate which catalyst is the best for crack-
ing or hydrotreating the light cycle oil (LCO) fraction, which is ob-
tained directly from fluid catalytic cracking units. A set of zeolites
with medium-sized (ZSM-5, MCM-22, ITQ-2), large (USY, Beta),
quality of diesel fuels and is equivalent to the percentage
and ultralarge pores (UTD-1), as well as a mesoporous MCM-41,
were used as catalysts at 723 K. The results demonstrate that pore
size and topology have a strong influence on diffusion, and conse-
quently, on activity and selectivity in reactions such as ring opening,
assigned a cetane number of 100; heptamethylnonane has
a long delay period and has a cetane number of 15. Just
as the octane number is meaningful for automobile fuels,
the cetane number is a means of determining the ignition
by volume of cetane, in the blend with heptamethylnonane,
that matches the ignition quality of the test fuel. Therefore,
the direct introduction of LCO into the diesel fuel low-
dealkylation, transalkylation, hydride transfer, and coke formation. ers the overall quality and can limit its use unless LCO is
UTD-1 generally has the highest activity per framework Al owing to previously hydrotreated. Therefore, the trend in LCO hy-
the pore size and topology of this zeolite that enables flat molecules drotreating is not only to decrease the sulfur content, but
to diffuse easily inside the pores. According to the results, zeolites
with medium-sized pores are adequate in combination with large-
pore zeolites to crack naphthenes and fused aromatic–naphthenic
rings, of the type present in LCO, to produce propene. Large-pore
number increases, as can be deduced from the cetane index
zeolites show good selectivity for naphthenic ring opening and
also to partially hydrogenate the polyaromatic compounds
to form aromatic–naphthenic molecules that can then be
cracked to alkylaromaticcompounds. In thisway, the cetane
(
CI) of the compounds produced.
appear to be better suited for hydrotreating LCO. Beta zeolite
is a catalyst that is especially suitable for both processes.
�c 2001
Academic Press
Key Words: decalin cracking; tetralin cracking; light cycle oil
cracking
INTRODUCTION
Vacuum gasoil and residual oil are cracked in fluid cata-
lytic cracking (FCC) units to liquid fuels (gasoline and light
cycle oil, LCO) for automotive fuels and short-chain olefins
that are used as feed stocks in petrochemistry. While FCC
gasoline has acceptable research octane (RON) and mo-
tor octane numbers (MON), the LCO produced has a low
cetane index (� 20) and a high content of aromatics and
sulfur. The ability of diesel fuel to burn with the proper
characteristics is described by its cetane number, a mea-
sure of ignition delay. Excessively long ignition delays (low
In this process, a moderate amount of H2 is consumed,
while the cetane index increases. On the other hand, when
the demand for LCO is not large, as opposed to that for
C3 and C4 olefins, LCO can be reintroduced into the FCC
unit to crack the naphthenes, aromatic naphthenes, and the
alkyl chains in the alkylaromatic molecules.
Therefore, regardless of whether hydrotreating or re-
cracking of LCO is chosen, the cracking of naphthenes and
aromatic naphthenes must be carried out. However, de-
pending on the method chosen, the naphthene ring can be
opened or the cracking process can be used to produce a
large amount of small olefinic molecules.
1
To whom correspondence should be addressed. Fax: 34 96 3877809.
E-mail: acorma@itq.upv.es.
34
0021-9517/01 $35.00
Copyright �c 2001 by Academic Press
All rights of reproduction in any form reserved.

LIGHT CYCLE OIL CATALYTIC PROCESSING
35
We have studied the two methods by selecting two and the selected particle size was 0.50 to 0.75 mm, for which
molecules, decalin and tetralin, as model reactants. Once no control of internal diffusion was observed. The exper-
the catalytic cracking pattern for both molecules has been imental procedure used in this work has been described
determined, one can rationalize the influence of the pore previously (6).
topology of the zeolite on product distribution and, in this
Reaction products were analyzed by gas chromatogra-
way, select the catalyst that is most suitable for performing phy (GC) in a Hewlett Packard 6890 Series chromatograph
naphthenic ring opening or deeper cracking to short-chain using an HP-1 capillary column (100 m long) connected to
olefins.
a thermal conductivity detector and to a flame ionization
detector in series.
EXPERIMENTAL
Cumulative average conversion (X) is defined as the total
number of carbon atoms of outlet hydrocarbons, other than
the feed, divided by the number of carbon atoms in the
reactant fed. The product yield is defined as the moles of
product per mole of reactant fed.
Thermalcrackingexperimentswere carried out with both
reactants. Allthe reaction conditionswere the same asthose
in the presence of the catalyst. The conversions due to ther-
mal cracking at the longest TOS were practically negligible,
i.e., 1.0% for tetralin and 0.3% for decalin. In any case, these
thermal results were taken into account when determining
conversion and product yields.
Materials
The USY zeolite was prepared from a partially
ammonium-exchanged NaY zeolite (SK-40) by repeated
steam calcination and ammonium exchange. The unit cell
size of the resultant sample was 24.34 A˚ . A UTD-1 zeolite
was synthesized from a borosilicate, which was prepared in
the Alform accordingto the procedure described elsewhere
(
1), by substituting boron for aluminum. The ITQ-2 zeolite
was obtained by swelling an MCM-22 (Si/Al = 15) precur-
sor with hexadecyltrimethylammonium bromide followed
by delamination. Details of this procedure are described
elsewhere (2).
Beta, MCM-22 (3), and ZSM-5 (4) were synthesized ac-
cording to the procedures reported in the literature. An
MCM-41 sample with an Si/Al ratio of 17 was synthesized
according to the procedure described in Ref. (5) using a
hexadecyltrimethylammonium (Panreac 98 wt% ) cation as
a template and silica (Aerosil Degussa), pseudoboehmite
RESULTS AND DISCUSSION
The dimensions of decalin and tetralin as determined by
the Cerius 2 Molecular Simulations Inc. program are given
in Fig. 1. The distances are given in angstroms, and the van
der Waals diameter of hydrogen should be added. Even
with the approximations made, it can be seen that these
molecules will easily penetrate the 12-member-ring (MR)
zeolite pores, but the diffusion will be hindered in the case
of 10 MR zeolites. The same was assumed for molecules
formed by three condensed rings. A series of zeolites with
(
Catapal B, Vista), and sodium aluminate (Carlo Erba) as
the aluminum sources. The sample was activated at 813 K.
The size of the crystallites in the zeolite was determined
from the SEM results. Surface area and pore volume were
measured by nitrogen adsorption according to the BET
method.
1
4
0, 12, and 14 MR pores as well as a mesoporous Al-MCM-
1 were then prepared. The physicochemical characteristics
are given in Table 1. We selected two 10 MR pore zeo-
lites, which perform well in the production of short-chain
Catalytic Experiments
Catalytic experiments were performed at 723 K and at- olefins (MCM-22 (3) and ZSM-5 (7)), as potential catalysts
mospheric pressure in a fixed-bed glass tubular reactor together with ITQ-2 (3), two zeolites with large pores (USY,
(
300 mm � 16 mm i.d.) with an independently controlled Beta), and a zeolite with ultralarge pores (UTD-1).
three-zone heater. The catalyst-to-oil ratio, defined as the Conversion plots obtained for decalin and tetralin at dif-
amount of catalyst divided by the total amount of reactant ferent contact times, at a fixed time on stream (45 s for
(
0
tetralin or decalin) fed in each experiment, wasvaried from decalin and 49 s for tetralin), are given in Fig. 2. The slopes
.005 to 0.100 g � g , depending on the catalyst or reactant in the conversion vs contact time curves, when the contact
�
1
used, in order to keep the same range ofconversion. The dif- time is close to zero, were obtained. They represent the ini-
ferent catalyst-to-oil ratios were obtained by changing the tial reaction rates (Table 2). From the initial reaction rates
amount of catalyst and feeding a constant weight of reac- and the turnover numbers (TON) (calculated by dividing
tant (2 g). In order to obtain different conversion levels for the initial reaction rates by the number of micromoles of
a fixed catalyst-to-oil ratio, the final time on stream (TOS), pyridine per gram of catalyst which are still adsorbed at
�
i.e., the time the catalyst has been working, was varied by 400 C desorption temperature and which correspond to
changing the reactant feed rate with a positive displacement the number of strongest acid sites), it is clear that those
pump. The times of the reaction were 49, 121, 243, 487, and values are higher for large-pore than for medium-pore ze-
1
214 s for decalin cracking and 45, 112, 224, 449, and 1120 s olites. Consequently, we conclude that decalin and tetralin
for tetralin cracking. Zeolites were pelletized and crushed, can penetrate some of the pores of the 10 MR zeolites,

´
´
3
6
CORMA, GONZALEZ-ALFARO, AND ORCHILLES
FIG. 1. Dimensions of decalin and tetralin molecules and the fitting of two molecules in the channels of UTD-1.
but they do not reach all the acid sites. It is remarkable directional pore topology, gives the highest TON owing
that, when the MCM-22 was delaminated to give the corre- to the shape of the pores (7.5 � 10 A˚ ), which probably
sponding ITQ-2 sample, the accessibility of the acid sites to enables a fast diffusion of the relatively flat decalin and
the reactants improved considerably, the TON increasing tetralin molecules. The results of this study agree with
by about 25 for the reaction of decalin and tetralin.
those obtained during the catalytic cracking of another flat
However, even more surprising are the large differences molecule (triisopropylbenzene) in several 12 MR zeolites
in the intrinsic activity of the 12 MR zeolites. Despite and UTD-1 (1). In that study, it was shown that UTD-1 has
being crystallized with small crystallites (0.2 �m), Beta a higher intrinsic activity. It was hypothesized that this ze-
gives the lowest conversion. This is due to the fact that olite is a very adequate catalyst in reactions with larger flat
the Beta zeolite has two channels 7.2 � 6.2 A˚ in diame- molecules.
ter, and a third channel that is only 5.5 � 5.5 A˚ wide, so
In order to gain further insight into the interactions be-
that decalin and tetralin probably cannot diffuse in that tween the zeolite and the reactants, docking of decalin
direction. On the other hand, UTD-1, which has a uni- in Beta, USY, and UTD-1 zeolites was carried out with
TABLE 1
Properties of Catalysts Used in This Work
a
Acidity (�mol Py/g)
Brønsted (1545 cm 1)
�
Lewis (1450 cm
� 1
)
Area BET
(m /g)
Crystal size
Micropore volume
2
3
�
�
�
�
�
�
Catalyst
ZSM-5
MCM-22
ITQ-2
Beta
USY
Si/Al ratio
(�m)
(cm /g)
250 C
350 C
400 C
250 C
350 C
400 C
15
15
15
16
380
400
600
600
1.5–2.5
0.38
0.12
0.17
44
42
16
22
44
8
26
33
5
9
16
3
12
24
1
5
5
9
21
16
37
19
7
6
20
17
35
13
5
5
18
—
33
9
<0.2
0.6
0.2–0.3
0.24
0.161
0.091
b
15
UTD-1
A1-MCM-41
80
17
305
1
3
5
24
12
6
47
34
a
Acidity measured by pyridine adsorbed and desorbed at different temperatures according to Emeis et al. (17).
Framework Si/Al ratio from XRD data and Fichtner–Schmittler’s equation (18); Si/Al ratio of other zeolites based on chemical analysis.
b

LIGHT CYCLE OIL CATALYTIC PROCESSING
37
TABLE 2
the Cerius program. The results indicate that decalin and
tetralin diffuse within the pores of the three zeolites, but
theydiffuse much better in the poresofUTD-1. It isremark-
able that, in the case of the 14 MR zeolite, two molecules
fit together in the channel (see Fig. 1). It is thus concluded
from the theoretical as well as from the experimental work
that UTD-1 reacts especially well with flat molecules that
can diffuse through its pore structure. This zeolite is, thus,
better than the others with a tridirectional large-pore net-
work system.
The mesoporous MCM-41 material with a pore diameter
of 35 A˚ is almost inactive due to the small number of acid
sites that are strong enough for the cracking of the naph-
thenic ring. A similar conclusion was reached after cracking
n-alkanes with MCM-41 (8).
Rates and TON in Decalin and Tetralin Cracking over Different
Catalysts Measured at 49 and 45 s of TOS for Tetralin and Decalin
respectively at 723 K
Rates (g � g� � h� 1
1
)
TON400�C (g � �mol Py� 1 � h� 1
)
Catalyst
ZSM-5
MCM-22
ITQ-2
Beta
USY
UTD-1
Al-MCM-41
Tetralin
Decalin
Tetralin
Decalin
106
39
41
395
561
410
10
103
60
58
8.8
1.6
41.4
79
8.6
2.5
58.0
90.8
180
714
4.7
454
899
714
14
112
410
3.3
Product Distribution
Tetralin cracking. A large number of products is ob-
tained after reacting tetralin; 40 major products were de-
tected and quantified. The yield of each product was plot-
ted against total conversion, and an example of the curves
obtained for some products with the ZSM-5 catalyst are
given in Fig. 3. From these curves, the initial selectivities
were calculated, and the results are given in Table 3. When
reacting tetralin, C1 to C4 fraction, naphthalene, benzene,
and C10 olefin/naphthene aromatics (O/N-AC10) (mainly
methylindanes) account for close to 85% of the total con-
version, and their relative initial selectivity changes from
one zeolite to another.
It is quite clear that, to form the above products
from tetralin, reactions such as ring opening, dealkylation,
transalkylation, and hydrogen transfer must be involved.
It has been reported in the literature (9) that the reaction
of tetralin can be initiated by two different mechanisms: a
bimolecular mechanism, which is responsible for the for-
mation of species related to phenylbutyl tetralin (PBT),
and a unimolecular mechanism. Under our reaction condi-
tions and a relatively high temperature, the amount of PBT
was negligible and, consequently, the unimolecular initia-
tion step (Scheme 1), involving the attack of a Brønsted
site either on the aromatic ring (10) or on the naphthenic
ring (11, 12), dominates. �-Scission and protolytic cracking
of the resultant carbenium and carbonium ions can explain
the opening of the naphthenic ring, forming alkenyl and
alkylbenzenes respectively (Scheme 1). On the other hand,
an isomerization reaction can occur by means of a PCP
mechanism (13), which may be responsible for the contrac-
tion of the ring and the formation of methylindanes and
methylindenes.
Dealkylation reactions (14) produce gaseous molecules
and aromatics with fewer than 10 carbon atoms (Tables 3
and 4), while disproportionation reactions between carbe-
FIG. 2. Cumulative conversion versus contact time (weight hourly
space velocity ) for different catalysts in tetralin and decalin cracking.
�
1
Experimental points: (᭹) UTD-1, (�) USY, (᭝) Beta, (᭡) MCM-22, (᭛) nium ions and aliphatic species (15) or between aromat-
ITQ-2, (᭿) ZSM-5, and (ᮀ) Al-MCM-41 catalysts.
ics and olefins can explain the formation of aliphatic and

´
´
3
8
CORMA, GONZALEZ-ALFARO, AND ORCHILLES
FIG. 3. Yield (mol% ) versus conversion of primary products in tetralin cracking over ZSM-5 zeolite. Experimental points: (᭝) 0.0125, (ᮀ) 0.0250,
and (᭺) 0.1000 catalyst/oil ratios. Unbroken lines represent catalyst behavior without deactivation.
aromatic compounds with more than 4 and 10 carbon atoms (Table 4). Scheme 1 shows the main reactions that occur
respectively. Scheme 2 shows, for example, transalkylation during the cracking of tetralin and the products involved.
reactions between several aromatic species and the butane The fact that both unimolecular and bimolecular reactions
and butene obtained from tetralin by ring opening (10).
take place and that intermediates and final products of dif-
The formation of naphthalene as a primary product is ferent dimensions are formed precludes the pore dimen-
explained by fast hydrogen transfer between tetralin and sions of the zeolite affecting the initial selectivities.
the adsorbed carbenium ions formed during the isomeriza-
The resultsin Tables3and 4clearlyshowthat bimolecular
tion and ring opening of tetralin. Saturation of the ben- hydrogen transfer (naphthalene and P/O ratio) occurs to a
zene ring of tetralin was detectable only with zeolite USY much smaller extent with the 10 MR ZSM-5 zeolite than
and to a much lesser extent than the amount of naphtha- with the other molecular sieves; hydrogen transfer probably
lene that was formed. The extent of the bimolecular hy- occurs mainly on the outer surface of the crystals of ZSM-
drogen transfer reactions can also be evaluated by means 5. MCM-22 results in greater hydrogen transfer than does
of the paraffin/olefin (P/O) ratio in the reaction products ZSM-5, probably due to the larger external surface of the

LIGHT CYCLE OIL CATALYTIC PROCESSING
39
TABLE 3
Initial Molar Selectivities of Products in Tetralin Cracking at 723 K
Catalyst
Product
ZSM-5
MCM-22
ITQ-2
Beta
USY
USY-24.26
UTD-1
Al-MCM-41
C1 + C2 Fractions
Propane
Propene
0.0017
0.0209
0.3534
0.0690
0.0120
0.0040
0.0059
0.5130
0.0035
0.0007
0.0721
0.0000
0.0243
0.0000
0.1498
0.1118
0.0000
0.1008
0.0550
0.0089
0.0037
0.0206
0.0000
0.0498
0.1520
0.0612
0.0058
0.0000
0.0046
0.1567
0.0000
0.0035
0.0457
0.0033
0.0141
0.0079
0.3373
0.0415
0.0000
0.2003
0.0649
0.0000
0.0034
0.0170
0.0049
0.0400
0.1089
0.1225
0.0167
0.0059
0.0088
0.1951
0.0019
0.0033
0.0592
0.0000
0.0216
0.0276
0.3247
0.0112
0.0000
0.2417
0.0431
0.0114
0.0037
0.0350
0.0092
0.0398
0.1639
0.0668
0.0093
0.0033
0.0043
0.1753
0.0000
0.0010
0.0610
0.0150
0.0484
0.0162
0.3153
0.0225
0.0000
0.1722
0.0942
0.0252
0.0063
0.0290
0.0024
0.0453
0.0663
0.1286
0.0266
0.0074
0.0164
0.0933
0.0173
0.0013
0.0522
0.0125
0.0316
0.0635
0.1119
0.0123
0.0219
0.4332
0.0371
0.0134
0.0035
0.0369
0.0000
0.0454
0.0830
0.1363
0.0160
0.0048
0.0166
0.1076
0.0208
0.0042
0.0758
0.0000
0.0000
0.1084
0.1710
0.0090
0.0000
0.3503
0.0575
0.0118
0.0015
0.0267
0.0000
0.0412
0.0310
0.0934
0.0138
0.0059
0.0203
0.0807
0.0054
0.0042
0.0201
0.0000
0.0109
0.1586
0.2338
0.0062
0.0000
0.3461
0.0319
0.0088
0.0012
0.0459
0.0016
0.0000
0.0370
0.0283
0.0047
0.0051
0.0185
0.0670
0.0000
0.0000
0.0000
0.0000
0.0155
0.0145
0.4563
0.0968
0.0000
0.1845
0.0331
0.0110
0.0007
0.1053
C4 Fraction
C5 Fraction
C6 Aliphatics
C6 Naphthenics
a
AC6
C7 Aliphatics
C7 Naphthenics
a
a
AC7 + AC8
a
A-AC9
O/N-AC9
A-AC10
a
a
a
O/N-AC10
DO/NO-AC10a
c-Decalin
Naphthalene
O/N-AC11a
Methylnaphthalenes
O/N-AC12a
Coke
aAC6, C6 aromatic (C6H6); AC7, C7 aromatic (C7H8); AC8, C8 aromatic (C8H10); A-AC9, C9 alkyl aromatics (C9H12); O/N-AC9, C9 olefin
or naphtheno aromatics (C9H10); A-AC10, C10 alkyl aromatics (C10H14); O/N-AC10, C10 olefin or naphtheno aromatics (C10H12); DO/NO-
AC10, C10 diolefin or naphtheno-olefinic aromatics (C10H10); O/N-AC11, C11 olefin or naphtheno aromatics (C11H14); O/N-AC12, C12 olefin
or naphtheno aromatic (C12H16). Examples of these molecules are shown in Schemes 1–4.
SCHEME 1. Reaction scheme for catalytic cracking of tetralin. See Table 3 for symbols.

´ ´
CORMA, GONZALEZ-ALFARO, AND ORCHILLES
4
0
TABLE 4
which has not undergone dealkylation. Moreover, due to
the instability of methylcarbenium ions, alkyl C9 aromatics
cannot form by cracking of tetralin and the low selectivity
of methane over all the catalysts shows that a minimum
number of olefinic or naphthenic C9 aromatics is produced
through this route. Thus, C9 aromatics must be produced
through disproportionation reactions. On the other hand,
coke selectivity (Table 3) is expressed as a hypothetical
molecular formula of C10H8 and was based on the exper-
imental H/C ratio of coke (0.8). Coke clearly has a more
complex structure, and a higher molecular weight, in which
polyaromatic condensed rings exist, but we assume that
coke aggregates and their selectivity increase as tetralin
molecules become attached to the coke aggregates and de-
hydrogenate. If this is so, then the sum of the selectivities
of coke and aromatics with more than 10 carbon atoms is
a good indication of the molecules that not to dealkylate,
being equal to the difference between unity and the selec-
tivity to dealkylation. Thus, we can calculate the dealkyla-
tion selectivity during tetralin cracking from Eq. [1]. The
dealkylation selectivity values obtained from Eq. [1] and
the selectivities given in Table 3 are reported in Table 5.
Molar Selectivities to Gaseous Products and P/O Ratio Obtained
from the Reaction of Naphthenes
Gases
P/O
Catalyst
ZSM-5
MCM-22
ITQ-2
Beta
USY
USY-24.26
UTD-1
Al-MCM-41
Tetralin
Decalin
Tetralin
Decalin
0.4450
0.2630
0.2763
0.2797
0.2426
0.2647
0.1656
0.0669
1.0872
1.1540
0.9869
0.6606
0.2677
—
0.07
0.60
1.14
0.51
1.85
1.30
2.23
0.27
0.27
0.49
0.62
0.70
2.32
—
0.2734
0.8632
1.09
0.24
former due to the well-defined 12 MR “cups” of � 7 � 12 A˚
16). This is supported by the fact that, when the number
(
of those “cups” at the external surface increases after the
delamination process (2) that produces the ITQ-2 structure,
hydrogen transfer increases, even when the density of the
acid sites in ITQ-2 is half that in MCM-22 (Table 1).
In the case of ZSM-5 when the naphthenic ring is opened,
the long-chain alkyl or alkenyl group in the resultant alkyl-
benzene rapidly cracks, giving propene and benzene as the
major products.
Sdealkylation = 1 � Scoke � SC12 aromatic � SC11 aromatic
�
SC10 aromatic � SC9 aromatic.
[1]
For a more quantitative comparison of the selectivity
of dealkylation of the different zeolites, we calculated
Thus, we suggest that cracking LCO in the FCC unit,
the dealkylation selectivity of tetralin based on the reac- with a ZSM-5 additive in the catalyst, will crack the alkyl
tions represented by Schemes 1 and 2. According to these chains of the cyclic products present in LCO to give mainly
schemes, each aromatic molecule with 10 or more carbon propene, leavinghighlyrefractoryaromaticrings. However,
atoms is probably obtained from one tetralin molecule with aromatic–naphthenic rings, ZSM-5 will be less active
SCHEME 2. Possible disproportionation reactions (transalkylation + cracking) in tetralin cracking. See Table 3 for symbols.

LIGHT CYCLE OIL CATALYTIC PROCESSING
41
TABLE 5
for ring opening, which will occur mainly on the external
surface of the crystallites.
a
Molar Selectivities to Dealkylation Obtained
in Tetralin and Decalin Cracking
Thus, in a first approximation, a suitable catalyst for
cracking LCO, while producing propene, should involve
a catalytic function able to open the naphthenic rings to-
gether with ZSM-5 zeolite that can crack the branched
naphthenic or alkylaromatic molecules while producing
propene.
On the other hand, if the goal is to improve the qual-
ity of the LCO to be introduced into the diesel pool by
hydrotreating the LCO with a bifunctional catalyst based
on a zeolite, ZSM-5 or other medium-pore zeolites do not
Catalyst
ZSM-5
MCM-22
ITQ-2
Beta
USY
UTD-1
Al-MCM-41
Tetralin
Decalin
0.525
0.311
0.280
0.256
0.223
0.157
0.082
0.548
0.737
0.671
0.521
0.224
0.232
0.540
aCalculated from Eq. [1].
FIG. 4. Yield (mol% ) versus conversion for primary products in decalin cracking over ZSM-5 zeolite. Experimental points: (᭝) 0.0125 and (ᮀ)
0
.0250 catalyst/oil ratios. Solid lines represent catalyst behavior without deactivation.

´ ´
CORMA, GONZALEZ-ALFARO, AND ORCHILLES
4
2
seem to be suitable as an acid function, since they can hardly cracking point of view, Beta may be a good component of
deal with the polyaromatic and aromatic–naphthenic rings an acid catalyst, since it produces the smallest amount of
present in the LCO fraction. Moreover, after opening of naphthalene and coke of the large-pore zeolites studied.
the rings, the cracking of the alkyl or alkenyl chains takes It shows good selectivity to alkylaromatics or naphtheno-
place too rapidly, leading to too many gaseous molecules.
aromatics and the highest selectivity to propene. Thus, Beta
In the case of 12 and 14 MR zeolites, it is seen (Table 3) zeolite, in combination with ZSM-5, may be a very suitable
that USY and UTD-1 produce a large amount of naphtha- catalyst in the FCC unit for the cracking of LCO to propene.
lene and coke. It isremarkable that UTD-1, even with a very
Because of the characteristics of Beta zeolite outlined
high Si/Al ratio (80), is still highly selective for bimolecular above and its high selectivity to ring opening and low selec-
reactions. This is probably due to the unidirectional char- tivity to gaseous molecules, it may also be a suitable catalyst
acter and the size of the pores that enable the close fit of component for LCO partial hydrogenation and ring open-
two molecules at distances great enough for bimolecular ing.
reactions to occur (see Fig. 1).
In conclusion, when reacting aromatic–naphthenic con-
In order to differentiate between the influence of the densed rings as present in LCO, and if one only wants to
framework Si/Al ratio and the influence of the pore topol- open the ring in order to improve the cetane number, then
ogy, a second USY zeolite with a framework Si/Alratio of60 a large-pore zeolite is preferred over medium-pore zeolites.
(
(
unit cell size = 24.26 A˚ ) was also studied. The results show In that case USY or Beta would be preferred. On the other
Tables 3 and 4) that hydrogen transfer and coke formation hand, if LCO has to be cracked, to increase the yield of
decrease with the increasing ratio of the framework Si/Al propene, then a catalyst with a large-pore zeolite, in partic-
of the zeolite. Based on this and the selectivity obtained ular Beta and ZSM-5, would be more suitable.
with UTD-1, we suggest that the pore topology of UTD-1,
or even the presence of traces of cobalt, plays an important
role in the high hydrogen transfer and coke yields.
Decalin cracking. During the cracking of decalin,
gaseous molecules, C6 naphthenics (mainly methylcy-
Of the 12 MR zeolites, Beta is a good all-round zeolite as clopentane), and C10 aromatics (butylbenzenes) are pro-
far as activity, yield of gases, alkylaromatic formation, and duced in especially large amounts (Fig. 4 and Table 6).
hydrogen transfer are concerned. Indeed, from the LCO Moreover, the amount of tetralin and naphthalene formed
TABLE 6
Initial Molar Selectivities of the Products in Decalin Cracking at 723 K
Catalyst
Product
Methane
Ethene
Propane
ZSM-5
MCM-22
ITQ-2
Beta
USY
UTD-1
Al-MCM-41
0.0003
0.2199
0.0810
0.4336
0.3523
0.0817
0.0234
0.0559
0.0168
0.0183
0.0101
0.1407
0.1298
0.0320
0.2371
0.0215
0.0000
0.0099
0.0000
0.0019
0.0000
0.0000
0.0199
0.0000
0.1248
0.1136
0.3206
0.5950
0.1001
0.0575
0.1164
0.0176
0.0458
0.0556
0.1514
0.1045
0.0044
0.0642
0.0392
0.0252
0.0000
0.0032
0.0091
0.0009
0.0000
0.0125
0.0080
0.0733
0.0419
0.3088
0.5520
0.1252
0.0607
0.1208
0.0127
0.0403
0.0369
0.0900
0.1155
0.0185
0.1321
0.0257
0.0135
0.0000
0.0048
0.0045
0.0000
0.0000
0.0140
0.0037
0.0197
0.0220
0.1710
0.4434
0.1624
0.0651
0.1433
0.0090
0.0392
0.0325
0.0501
0.1527
0.0405
0.2076
0.0325
0.0148
0.0000
0.0007
0.0073
0.0023
0.0006
0.0200
0.0000
0.0046
0.0253
0.0431
0.1946
0.0573
0.0303
0.1017
0.0031
0.0317
0.0066
0.0275
0.2828
0.0585
0.3313
0.0347
0.0408
0.0000
0.0036
0.0071
0.0016
0.0000
0.0165
0.0006
0.0101
0.0072
0.0638
0.1915
0.0532
0.0318
0.1071
0.0020
0.0313
0.0190
0.0193
0.2403
0.0550
0.3765
0.0383
0.0156
0.0086
0.0046
0.0041
0.0033
0.0005
0.0209
0.0028
0.0327
0.0183
0.2962
0.5133
0.1072
0.0868
0.1756
0.0035
0.0259
0.0057
0.0258
0.1676
0.0465
0.1826
0.0198
0.0042
0.0000
0.0000
0.0000
0.0000
0.0000
0.0395
Propene
C4 Fraction
C5 Fraction
C6 Aliphatics
C6 Naphthenics
a
AC6
C7 Aliphatics
C7 Naphthenics
a
a
AC7 + AC8
a
A-AC9
O/N-AC9
a
a
A-AC10
O/N-AC10a
DO/NO-AC10a
Tetralin
Naphthalene
O/N-AC11a
Methylnaphthalenes
O/N-AC12a
Coke
aAbbreviations as in Table 3.

LIGHT CYCLE OIL CATALYTIC PROCESSING
43
SCHEME 3. Mechanisms of reactions of decalin over acid catalysts (capital letters as in Schemes 1 and 2).
is very small, indicating that the direct dehydrogenation of tive than 12 MR zeolites in reaction with decalin (Table
decalin by hydrogen transfer is negligible. Consequently, 2). This is also a clear indication that decalin hardly pene-
when reacting decalin, the ring is opened first, followed by trates the channels of the medium-pore zeolite. In the case
dealkylation and hydrogen transfer. Based on this and the of MCM-22, the reaction occurs on the external surface
nature of the rest of the products formed (Table 6), we can and, consequently, the activity per site increases strongly
draw the reaction scheme for decalin (Scheme 3). The reac- (Tables 1 and 2) when the external surface is increased
tion is initiated by the attack of a Brønsted site on a carbon– by delamination (ITQ-2 sample). In the case of the 12
carbon bond of the decalin (11) forming a carbonium ion, MR zeolites, the order of activity is the same as that pre-
which cracks protolytically to an alkylnaphthene carbe- viously observed when reacting tetralin; again the high
nium ion. The formation of gaseous molecules and naph- activity per site of the 14 MR UTD-1 zeolite is surp
thenes lighter than the feed can be explained by unimolecu- rising.
lar reactions, involving the isomerization and �-scission of
As far as selectivity is concerned, it is remarkable that no
the alkylnaphthene carbenium ion. Meanwhile, bimolec- alkylnaphthenes with longer chains are found. They readily
ular hydrogen transfer and hydride transfer produce C10 crack and transfer hydrogen, giving alkylaromatics. Thus,
alkyl, alkenyl, or naphthene aromatics. Other bimolecu- even ZSM-5 gives a reasonable number of alkylaromatics
lar disproportionation reactions, involving transalkylation but a smaller number than that produced with 12 MR zeo-
and cracking, are also responsible for the formation of lites. In the case of decalin, UTD-1 is again the most active
aromatics with more carbon atoms than the feed. Fur- catalyst in the bimolecular reactions and, more specifically,
thermore, C9 aromatics can be formed by a bimolecular hydrogen transfer. On the other hand, the dealkylation and
process such as that represented by Scheme 4. The naph- gas selectivities are higher for the 10 MR than for 12 MR
thenic species included in Scheme 4 can react further by hy- zeolites. Of the large-pore zeolites, Beta gives the greatest
dride and hydrogen transfer to give the corresponding aro number of gaseous molecules. This correlates well with the
matics.
pore dimensions of the zeolites, because the smaller the
pores, the greater the probability of secondary unimolec-
ular reactions such as dealkylation and cracking reactions.
The relatively large amounts of ethene produced by ZSM-
At this point by analyzing the rates of the reaction for de-
calin (Table 2), the selectivitiesfor dealkylation and gaseous
molecules (Tables 5 and 4), and the hydrogen transfer (P/O
ratio) (Table 4), we conclude the following: Decalin reacts
faster than tetralin, and 10 MR zeolites are much less ac-
5
and MCM-22 and the fact that Beta zeolite produces the
highest yield of propene of the 12 MR zeolites is notewor-
thy.
These results suggest that a good LCO cracking catalyst
for the production of naphthenes should involve a large-
pore zeolite for primary cracking. Beta is a very suitable
zeolite for this, together with ZSM-5 that can crack the
smaller products formed after the primary cracking on Beta
zeolite to give high yields of propene and ethene.
SCHEME 4. Possible reaction for formation of C9 species.

´
´
4
4
CORMA, GONZALEZ-ALFARO, AND ORCHILLES
CONCLUSIONS
ACKNOWLEDGMENTS
We thank the Spanish CICYT (project MAT97-1207-C03-01) for finan-
cial support.
Decalin and tetralin are excellent probe molecules for
studying the cracking behavior of cycloalkanes and aro-
matic naphthene condensed ring molecules in the LCO
fraction on either monofunctional or bifunctional zeolite-
based catalysts. Thus, studying the reactivity of these two
molecules on a series of 10, 12, and 14 MR zeolites showed
the following:
If the aim is to crack LCO in the FCC unit and to
maximize the amount of propene formed, a “cocktail” of
large- and medium-pore zeolites is preferred. The large-
pore zeolite (preferably Beta) will open the naphthenic
ring, while the 10 MR zeolite (ZSM-5 or MCM-22) will
crack the branched naphthenics or alkylaromatics formed
to propene.
If, on the other hand, the goal is to improve the qual-
ity of the LCO, i.e., to decrease the number of aromatics
and increase the CI, the LCO should be hydrotreated with
a bifunctional catalyst based on a 12 MR zeolite (USY or
even better Beta zeolite). The partially hydrogenated poly-
cyclic aromatic rings will then open and the cracking of the
resultant alkyl chains will be minimized.
In the case of the FCC operation, the aromatic naph-
thene condensed rings will crack on large-pore USY or
Beta zeolite, opening the naphthene ring and minimizing
the dealkylation of the alkylaromatic formed.
REFERENCES
1
2
. Mart ´ı nez-Triguero, J., D ´ı az-Caba n˜ as, M. J., Camblor, M. A., Forn e´ s,
V., Maesen, Th. L. M., and Corma, A., J. Catal. 182, 463 (1999).
. Corma, A., Forn e´ s, V., Pergher, S. B., Bouglass, J. G., and Maesen,
Th. L. M., Nature 396, 353 (1998); Corma, A., Forn e´ s, V., Mart ´ı nez-
Triguero, J., and Pergher, S. B., J. Catal. 186, 57 (1999).
3
4
. Corma, A., Corell, C., and P e´ rez-Pariente, J., Zeolites 15, 2 (1995).
. Szostack, R., “Molecular Sieves, Principles of Synthesis and Identifi-
cation,” Catalysis Series, Van Nordstrand Reinhold, New York, 1989.
. Beck, J. S., Chu, C. T. W., Johnson, I. D., Kresge, C. T., Leonowicz, M.
E., Roth, W. J., and Vartuli, J. C., Patent WO 91/11391, 1991.
. Corma, A., Davis, M., Forn e´ s, V., Gonz a´ lez-Alfaro, V., Lobo, R., and
Orchill e´ s, A. V., J. Catal. 167, 438 (1997).
5
6
7
8
9
. Degnan, T. F., Chitnis, G. K., and Schipper, P. H., Microporous Meso-
porous Mater. 35–36, 245 (2000).
. Corma, A., Grande, M. S., Gonz a´ lez-Alfaro, V., and Orchill e´ s, A. V.,
J. Catal. 159, 375 (1996).
. Sato, K., Iwata, Y., Miki, Y., and Shimada, H., J. Catal. 186, 45 (1999).
1
0. Mostad, H. B., Riis, T. U., and Ellestad, O. H., Appl. Catal. 63, 345
1990).
11. Abbot, J., and Wojciechowski, B. W., Can. J. Chem. Eng. 67, 833 (1989).
(
12. Corma, A., Mochol ı´ , F., Orchill e´ s, A. V., Koermer, G. S., and Madon,
R. J., Appl. Catal. 67, 307 (1991).
1
1
3. Weitkamp, J., Ind. Eng. Chem. Prod. Des. Dev. 21, 550 (1982).
4. Corma, A., Miguel, P. J., Orchill e´ s, A. V., and Koermer, G., J. Catal.
145, 181 (1994).
It is remarkable that, owing to its pore size and shape, 15. Zhao, Y. X., Bamwenda, G. R., Groten, W. A., and Wojciechowski, B.
W., J. Catal. 140, 243 (1993).
UTD-1 has higher activity per active site than any of the
1
6. Lawton, S. L., Leonowicz, M. E., Partridge, R. D., Chu, P., and Rubin,
M. K., Microporous Mesoporous Mater. 23, 109 (1998).
7. Emeis, C. A., J. Catal. 141, 347 (1993).
8. Fichtner-Schmittler, H., Lohse, U., Engelhardt, G., and Patzelova, V.,
Cryst. Res. Technol. 19, K1 (1984).
other zeolites studied and is highly selective for bimolec-
ular reactions (hydrogen transfer) owing to the possibility
of fitting two molecules that are in close proximity in the
channels.
1
1