Disproportionation of isopropylnaphthalene on zeolite catalysts

Article abstract of DOI:10.1016/S0021-9517(03)00237-9

Disproportionation of isopropylnaphthalene (IPN) was tested over H-mordenite, HY, H-beta zeolites and over amorphous aluminosilicate in the range of 150-300°C. High β,β-selectivity in diisopropylnaphthalene (DIPN) product obtained over zeolites was observed. However, the 2,6-DIPN/2,7-DIPN mole ratio was dependent on the pore structure of the applied zeolite and on the temperature. Over H-mordenites 2,6-DIPN was the most preferred isomer, whereas 2,7-DIPN was favored over HY and H-beta zeolites. Such disproportion in isomer predominating in the DIPN product can be explained by a bimolecular mechanism of disproportionation. Due to the zeolite pore architecture the bent transition-state complex, leading to 2,7-DIPN (HY and H-beta) or more linear, leading to 2,6-DIPN (H-mordenite), was preferred. At high temperatures the monomolecular disproportionation mechanism (dealkylation realkylation) dominated and concealed (simultaneously with side reactions) the shape-selectivity effect. As a result the 2,6-DIPN/2,7-DIPN mole ratio in the product approached equilibrium value.

Full text of DOI:10.1016/S0021-9517(03)00237-9

Journal of Catalysis 220 (2003) 13–22
www.elsevier.com/locate/jcat
Disproportionation of isopropylnaphthalene on zeolite catalysts
Robert Brzozowski ∗ and Wincenty Skupin´ski
Industrial Chemistry Research Institute, Rydygiera 8, 01-793 Warsaw, Poland
Received 3 December 2002; revised 6 March 2003; accepted 16 May 2003
Abstract
Disproportionation of isopropylnaphthalene (IPN) was tested over H-mordenite, HY, H-beta zeolites and over amorphous aluminosilicate
◦
in the range of 150–300 C. High β,β-selectivity in diisopropylnaphthalene (DIPN) product obtained over zeolites was observed. However,
the 2,6-DIPN/2,7-DIPN mole ratio was dependent on the pore structure of the applied zeolite and on the temperature. Over H-mordenites
2,6-DIPN was the most preferred isomer, whereas 2,7-DIPN was favored over HY and H-beta zeolites. Such disproportion in isomer pre-
dominating in the DIPN product can be explained by a bimolecular mechanism of disproportionation. Due to the zeolite pore architecture the
bent transition-state complex, leading to 2,7-DIPN (HY and H-beta) or more linear, leading to 2,6-DIPN (H-mordenite), was preferred. At
high temperatures the monomolecular disproportionation mechanism (dealkylation realkylation) dominated and concealed (simultaneously
with side reactions) the shape-selectivity effect. As a result the 2,6-DIPN/2,7-DIPN mole ratio in the product approached equilibrium value.
2003 Elsevier Inc. All rights reserved.
Keywords: Shape selectivity; Zeolite; Disproportionation; Isopropylnaphthalene; Diisopropylnaphthalene
1. Introduction
ily than the 2,7-DIPN through the mordenite channels. The
difference in diffusion between 2,6- and 2,7-DIPN was con-
firmed by DIPN sorption experiments [4,5]. According to
some authors [6,11,15,16] a difference in the critical diam-
eter between 2,6- and 2,7-DIPN molecules is too small to
cause such differences. However, Tasi et al. [4] recently cal-
culated the molecular dimensions of DIPN molecules and
obtained a significantly smaller cross section of 2,6-DIPN
than 2,7-DIPN.
Other authors [6,7,9] postulated that the activated com-
plex producing 2,6-DIPN is less hindered in the mordenite
pores than that for 2,7-DIPN. This supposition seems to be
confirmed by Kim et al.’s [9] results who found high 2,6-
DIPN/2,7-DIPN ratios both in the product encapsulated in
the zeolite pores and in the bulk product outside the pores.
Song and co-workers [11,15] calculated frontier elec-
tron density for electrophilic substitution at different carbon
atoms of the naphthalene nucleus and they found that for 2-
isopropylnaphthalene (2-IPN) molecules position 6 is kinet-
ically preferred to position 7. Therefore, the product formed
in the spatially hindered environment in a mordenite chan-
nel is more abundant in 2,6-DIPN than 2,7-DIPN. However,
this result is in contradiction with calculations of electrosta-
tic potential for 2-IPN by Tasi et al. [4] who did not confirm
distinguishing between positions 6 and 7 in electrophilic aro-
matic substitution.
2,6-Diisopropylnaphthalene (2,6-DIPN), isolated from
the isomeric mixture, can be used as a raw material for
the production of advanced polyester fibers, films, and plas-
tics such as thermotropic liquid crystalline polymers [1,2].
Although the most abundant isomers in a DIPN mix-
ture at thermodynamic equilibrium are 2,6- and 2,7-DIPN
(> 80%) [3–5], further improvement in 2,6-DIPN yield can
be achieved if the synthesis is carried out over wide pore
zeolites, such as mordenite, Y, beta, or L. Formation and
diffusion of bulky DIPN isomers with substituents at α po-
sitions of the naphthalene nucleus are restricted in zeolite
pores. Alkylation of naphthalene carried out over mordenite
catalysts has an additional advantage; apart from high β, β-
selectivity in diisopropylation, high 2,6-DIPN to 2,7-DIPN
ratios can be observed [5–14].
The reasons of higher yield of 2,6-DIPN than 2,7-DIPN
in the alkylation product are still widely discussed and they
are not quite clear. Horsley et al. [8] proved with the help
of molecular graphics and diffusion energy calculations that
the more linear 2,6-DIPN molecule can migrate more eas-
*
Corresponding author.
E-mail address: robert.brzozowski@ichp.pl (R. Brzozowski).
0021-9517/$ – see front matter
2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0021-9517(03)00237-9

14
R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
Regardless of the reason, shape-selectivity effects leading
Four zeolite catalysts in hydrogen form, HY, beta (HB),
and two mordenites, high-silica (HM-1) and low-silica (HM-
2), and amorphous aluminosilicate (AM) were tested in dis-
proportionation of IPN. The SiO2/Al2O3 mole ratios of the
catalysts were 5.8, 25, 84, 9.9, and 13.4, respectively. More
detailed characteristics of the catalysts, except for HM-2, are
described elsewhere [18]. The hydrogen form of mordenite
HM-2 (Na2O content 0.32 wt%) was prepared from the Na
form of Zeolon 900 (Norton Co.) by washing three times
with ammonia chloride solution according to the standard
procedure.
Disproportionation experiments were carried out in a
fixed-bed reactor (5 g of catalyst) at 0.8 MPa. Before each
experiment, the catalyst was activated at 450 ◦C in the flow
of air for at least 2 h. Subsequently the air stream was re-
placed for nitrogen and temperature was lowered to the ex-
perimental temperature. Reactant mixture was recirculated
through the catalyst bed at F/W = 70 g/(gcat h). Such flow
enabled the reactants to be one time passed through the cat-
alyst bed every 6 min. Samples of products were analyzed
by GC equipped with FID, using a 60-m-long HP-Innowax
capillary column. Details of the analytical method and the
DIPN isomers identification are described elsewhere [21].
To improve detection accuracy of DIPN isomeric composi-
tion, analyses of samples poor in DIPN were repeated using
increased amounts of injected sample and integrating only
the DIPN part of the GC spectra.
to high β-selectivity and high 2,6-DIPN/2,7-DIPN ratios
can be affected by side reactions occurring on the external
surface of zeolite and in the pore entrances. Alkylation of
naphthalene molecules in the pore entrances can lead to high
yields of 1,3-DIPN and 1,4-DIPN in the product [18]. Sec-
ondary side reactions such as disproportionation, transalky-
lation, and isomerization, occurring on the external surface
of the zeolite crystals, can also play an important role in the
synthesis of diisopropylnaphthalenes.
These side reactions can be suppressed by proper dealu-
mination of catalyst or by deactivation of external active sites
by deposition of compounds of silica, ceria, or other metal
[9–11,13,19,20]. Although β-selectivity and 2,6-DIPN yield
can be then improved, side reactions cannot be eliminated
entirely. Moreover, secondary reactions, such as isomeriza-
tion of 2,6- and 2,7-DIPN to one another [4] or dispropor-
tionation of 2-isopropylnaphthalene, are also possible in the
pores.
The influence of side reactions, always accompanying
alkylation, is frequently not considered. However, they can-
not be neglected in the alkylation environment, especially
when experiments are performed in a batch-type reactor and
reaction time is in the range of several hours [5–14].
The aim of the present work was to test disproportiona-
tion of IPN over shape-selective catalysts applied in naph-
thalene alkylation and to verify the possible influence of the
reaction on the alkylation product.
3. Results and discussion
2. Experimental
Results of disproportionation experiments carried out
over AM, HM-1, HM-2, HY, and HB catalysts are shown in
Tables 1–5, respectively. Results such as the IPN conversion
level, the 2-IPN content in monoisopropylnaphthalene (in-
dicating isomerization of IPN), the DIPN content in the re-
circulated mixture (indicating disproportionation progress),
and the isomeric composition of diisopropylnaphthalenes
Isopropylnaphthalene 99.7% pure (GC analysis) was
isolated by distillation from alkylates obtained previously.
Apart from 2-IPN (86.1% of total IPN) and 1-IPN it
contained small amounts of isopropyltetralines, ethylnaph-
thalenes, and other alkylnaphthalenes with molecular masses
of 184 and 198 (GC/MS).
Table 1
Results of IPN disproportionation over the AM catalyst
Temperature Time IPN conversion 2-IPN in IPNs DIPN content
Distribution of diisopropylnaphthalenes (mol%)
2,6-DIPN to
◦
( C)
150
(h)
(%)
(mol%)
(wt%)
2,6-DIPN 2,7-DIPN 1,3-DIPN 1,6-DIPN 1,7-DIPN Other 2,7-DIPN ratio
a
0.25
0.5
6
0.1
0.5
3.0
88.2
89.1
93.3
0.05
0.2
1.8
27.0
31.1
36.1
21.6
23.3
27.6
21.6
21.0
18.3
16.2
12.3
8.0
13.5
12.3
8.3
n.d.
n.d.
1.6
1.25
1.33
1.31
200
250
300
0.02
0.5
6
1.9
6.7
19.4
89.3
93.6
94.3
1.1
3.1
8.5
28.7
33.4
37.9
24.2
29.5
34.6
21.3
20.0
13.8
10.5
7.4
6.8
11.8
7.9
5.5
3.4
1.8
1.3
1.18
1.13
1.10
0.02
0.5
6
1.8
13.2
40.1
87.8
93.3
93.3
0.7
6.2
17.3
24.2
32.9
38.0
20.4
30.0
39.0
20.7
17.9
7.9
12.7
8.5
8.5
17.4
8.5
5.9
4.6
2.1
0.9
1.19
1.10
0.98
0.02
0.5
6
11.0
40.8
49.5
93.2
93.1
92.0
3.7
17.6
16.5
33.2
38.7
37.0
32.3
39.7
39.1
16.0
7.4
7.8
8.8
7.6
8.4
7.7
5.7
6.7
1.9
0.9
1.0
1.03
0.97
0.95
a
n.d., not detected.

R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
15
Table 2
Results of IPN disproportionation over the HM-1 catalyst
Temperature Time IPN conversion 2-IPN in IPNs DIPN content
Distribution of diisopropylnaphthalenes (mol%)
2,6-DIPN to
◦
( C)
150
(h)
(%)
(mol%)
(wt%)
2,6-DIPN 2,7-DIPN 1,3-DIPN 1,6-DIPN 1,7-DIPN Other 2,7-DIPN ratio
0.02
0.5
6
0.2
2.3
5.0
86.3
89.0
93.3
0.05
1.1
2.9
36.4
41.3
43.6
31.0
31.5
31.3
3.2
3.7
4.4
10.2
8.5
7.4
13.9
11.1
10.3
5.3
3.9
3.1
1.17
1.31
1.39
200
250
300
0.02
0.5
6
1.0
4.5
19.3
86.6
89.7
93.7
0.5
2.5
10.3
44.1
47.4
48.4
32.4
30.4
33.2
6.8
6.2
5.3
7.2
7.5
7.5
7.4
6.6
4.7
2.1
1.8
0.9
1.36
1.56
1.45
0.02
0.5
6
0.7
5.2
26.4
86.3
88.5
92.9
0.3
2.6
12.2
48.9
48.0
44.5
26.4
28.8
34.2
6.8
6.9
6.5
8.6
8.2
8.0
9.3
6.5
5.7
n.d.
1.6
1.1
1.85
1.67
1.30
0.02
0.5
6
2.5
18.2
48.2
87.0
90.7
91.9
1.2
7.9
17.5
49.3
43.7
37.8
28.3
32.5
38.8
5.4
6.8
7.5
8.5
8.9
7.7
6.9
6.9
7.3
1.5
0.6
0.6
1.74
1.35
0.97
Table 3
Results of IPN disproportionation over the HM-2 catalyst
Temperature Time IPN conversion 2-IPN in IPNs DIPN content
Distribution of diisopropylnaphthalenes (mol%)
2,6-DIPN to
◦
( C)
200
(h)
(%)
(mol%)
(wt%)
2,6-DIPN 2,7-DIPN 1,3-DIPN 1,6-DIPN 1,7-DIPN Other 2,7-DIPN ratio
0.5
6
0.1
0.5
86.0
86.3
0.05
0.3
39.3
45.3
29.4
30.4
12.8
7.7
9.6
10.0
8.9
6.7
n.d.
n.d.
1.34
1.49
250
0.02
0.5
6
0.7
2.6
10.9
86.6
88.5
92.2
0.4
1.5
5.5
48.1
53.9
50.2
26.0
25.7
29.3
9.2
5.4
5.3
9.0
8.2
9.1
7.7
4.9
4.7
n.d.
1.9
1.4
1.85
2.10
1.71
Table 4
Results of IPN disproportionation over HY zeolite
Temperature Time IPN conversion 2-IPN in IPNs DIPN content
Distribution of diisopropylnaphthalenes (mol%)
2,6-DIPN to
◦
( C)
150
(h)
(%)
(mol%)
(wt%)
2,6-DIPN 2,7-DIPN 1,3-DIPN 1,6-DIPN 1,7-DIPN Other 2,7-DIPN ratio
0.5
6
0.0
0.3
86.1
87.4
0.0
0.2
35.6
28.4
43.7
33.8
5.7
4.4
8.0
10.9
6.9
15.3
n.d.
7.3
0.82
0.84
200
250
300
0.02
0.5
6
1.4
4.0
25.8
87.9
93.2
94.6
0.6
2.4
14.4
21.9
23.6
36.4
30.0
38.0
47.5
3.5
3.1
3.7
15.3
12.9
6.5
21.4
16.2
5.0
7.8
6.2
1.0
0.73
0.62
0.77
0.02
0.5
6
3.0
11.2
39.8
88.1
93.5
93.8
1.4
5.6
21.1
29.5
29.7
37.3
39.6
43.1
45.2
3.9
3.1
4.7
11.2
10.3
6.9
12.2
11.1
5.2
3.6
2.7
0.6
0.74
0.69
0.83
0.02
0.5
6
5.4
39.2
53.6
89.7
92.5
93.0
2.9
19.6
20.7
18.2
37.0
41.1
23.9
43.0
44.9
3.8
4.8
4.8
15.6
6.9
4.8
29.9
7.1
4.0
8.6
1.5
0.4
0.76
0.86
0.92
Table 5
Results of isopropylnaphthalene disproportionation over HB catalyst
Temperature Time IPN conversion 2-IPN in IPNs DIPN content
Distribution of diisopropylnaphthalenes (mol%)
2,6-DIPN to
◦
( C)
150
(h)
(%)
(mol%)
(wt%)
2,6-DIPN 2,7-DIPN 1,3-DIPN 1,6-DIPN 1,7-DIPN Other 2,7-DIPN ratio
0.25
0.5
6
0.0
0.3
1.7
88.1
89.2
94.1
0.0
0.2
1.0
31.4
26.9
22.9
62.7
57.1
66.7
n.d.
4.8
2.5
5.9
6.3
5.0
n.d.
4.8
2.9
n.d.
n.d.
n.d.
0.50
0.47
0.34
200
250
0.02
0.5
6
1.7
9.2
29.1
89.4
94.1
94.6
1.0
5.3
12.8
25.1
27.2
31.0
58.8
58.2
56.0
4.4
3.8
2.7
6.3
6.2
5.2
5.4
4.5
5.0
n.d.
0.1
n.d.
0.43
0.47
0.55
0.02
0.5
6
9.8
35.8
48.1
91.4
93.4
93.9
4.4
19.2
19.1
33.6
37.5
38.3
48.1
45.6
42.3
4.4
4.0
5.9
8.0
7.2
7.7
5.2
5.3
5.3
0.7
0.5
0.5
0.70
0.82
0.91

16
R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
◦
Fig. 1. DIPN isomer distribution in disproportionation product obtained over amorphous aluminosilicate at 200 C.
◦
Fig. 2. DIPN isomer distribution in disproportionation product obtained over amorphous aluminosilicate at 250 C.
are compared. For better illustration of transformations oc-
curring in the reaction system the most characteristic time
dependencies of DIPN isomer concentrations are plotted in
Figs. 1–5, where results of experiments carried out over AM,
HM-1, and HY catalysts at 200 and 250 ◦C are shown.
As can be seen from Tables 1–5 and Figs. 1–5, for all
catalysts, the 2-IPN content in IPN increased with reaction
time, and the higher the temperature the more quickly the
isomeric composition of IPNs approached equilibrium con-
centrations.
Also IPN conversion increased with reaction time and
the increase was accelerated at elevated temperatures. At
the same time, the content of DIPN, naphthalene, triisopro-
pylnaphthalenes, and other alkylnaphthalenes in the recircu-
lated mixture increased. They were mainly produced by dis-
proportionation/transalkylation reactions, although dealky-

R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
17
◦
Fig. 3. DIPN isomer distribution in disproportionation product obtained over HM-1 catalyst at 200 C.
◦
Fig. 4. DIPN isomer distribution in disproportionation product obtained over HM-1 catalyst at 250 C.
lation and other side reactions were also observed. Espe-
cially at high temperature (300 ◦C) noticeable quantities of
propylene and other nonaromatic hydrocarbons were de-
tected in the products. Moreover, other by-products, such
as alkylnaphthalenes with other substituents than isopropyl,
tetraline, and binaphthyl derivatives, etc. were detected in
the products.
ments carried out over the AM catalyst at 200 and 250 ◦C,
respectively, concentrations of β, β-isomers (2,6- and 2,7-
DIPN) in diisopropylnaphthalene product increased with re-
action time and they were the most abundant isomers. Ini-
tially the product was rich in α, β-isomers (1,3-, 1,6-, and
1,7-) but they isomerized to the 2,6- and 2,7-DIPN. Dis-
tribution of other DIPN isomers (1,4-, 1,5-, and 2,3-) also
decreased with reaction time. At high temperatures changes
in isomeric composition were more rapid.
As can be seen from Figs. 1 and 2 which show the time
dependence of the DIPN isomeric composition in experi-

18
R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
◦
Fig. 5. DIPN isomer distribution in disproportionation product obtained over HY zeolite at 200 C.
Whereas intramolecular isomerization of 1,6-DIPN to
1.74 to 0.97. High 2,6-DIPN/2,7-DIPN ratios were also ob-
served over low-silica mordenite HM-2 (Table 3).
2,6-DIPN and 1,7-DIPN to 2,7-DIPN is possible, the 1,3-
DIPN depletion (simultaneously with 1,4- and 2,3-DIPN)
indicates intermolecular isomerization by disproportiona-
tion/transalkylation or dealkylation/realkylation-type mech-
anisms.
Although a 2,6-DIPN/2,7-DIPN ratio higher than 1 can
be expected by kinetics (vide AM catalyst), such signif-
icant difference in yield of 2,6-DIPN and 2,7-DIPN as
that observed over mordenite catalysts indicates shape-
selectivity effects. Also, a very low concentration of bulky
α, β-isomers, especially 1,3-DIPN, can be attributed to the
shape selectivity. From a comparison of the results in Ta-
bles 1 and 2 it is evident that the product obtained over
the AM catalyst was rich in 1,3-DIPN and its concentra-
tion decreased significantly with contact time, whereas the
DIPN obtained over HM-1 zeolite was poor in 1,3-isomer,
independent of time and temperature. Other bulky isomers
(1,6-DIPN and 1,7-DIPN) are also less abundant in the prod-
uct obtained over the HM-1 than over the AM catalyst.
A different relationship of DIPN isomer concentrations
with contact time was observed for zeolite HY (Fig. 5 and
Table 4). Although again β, β-isomers were the most abun-
dant in the DIPN mixture, the relation in β, β-isomer con-
tents was inverse to that obtained over HM-1, HM-2, and
AM catalysts; i.e., more 2,7-DIPN was formed than 2,6-
DIPN. The 2,6-DIPN/2,7-DIPN mole ratio was in the range
0.6–0.9 and preferentially approached the equilibrium value
while contacting with the catalyst.
In Fig. 2 isomerization of 2,6-DIPN to 2,7-DIPN at
250 ◦C can be observed. Initially DIPN mixtures contained
slightly more 2,6-DIPN than 2,7-DIPN but after 6 h their dis-
tribution was reversed. It can be deduced that 2,6-DIPN is
kinetically preferred but during further contact with catalyst
the mixture composition approaches equilibrium. At thermo-
dynamic equilibrium at 250 ◦C the distribution of 2,7-DIPN
is slightly higher than that of 2,6-DIPN [3]. Therefore, the
2,6-DIPN/2,7-DIPN ratio initially in the range of 1.2–1.3
(see Table 1) decreased with reaction time to 0.95–0.98.
The product of IPN disproportionation over the HM-1
catalyst (Figs. 3 and 4) contained DIPN isomers in de-
scending order: 2,6-DIPN > 2,7-DIPN > 1,3-DIPN ≈ 1,6-
DIPN ≈ 1,7-DIPN > other isomers. The distribution of
DIPN isomers did not change significantly during the re-
action. Similar steady time trends for DIPN concentrations
were observed by Sugi and Toba [7] in the alkylation of
naphthalene with propylene carried out over mordenite cata-
lysts.
The difference in concentrations of 2,6-DIPN and 2,7-
DIPN at 250 ◦C decreased with reaction time, slowly ap-
proaching equilibrium distributions. The initially high 2,6-
DIPN/2,7-DIPN ratio (Table 2) of 1.85 decreased to 1.3 due
to isomerization and other side reactions. At elevated tem-
peratures the decrease was more apparent and at 300 ◦C the
2,6-DIPN/2,7-DIPNratio decreased with reaction time from
Similarly as for HM-1 catalyst, the formation of 1,3-
DIPN was hindered over HY zeolite and its content in
the DIPN mixture did not exceed 6%, indicating a shape-
selectivity effect. However, 1,6-DIPN and 1,7-DIPN were
initially formed over the HY catalyst in quite high concentra-
tions; i.e., comparable to those observed over AM catalyst.
Also, other isomers were quite abundant in DIPN mixtures.

R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
19
All these bulky DIPNs isomerized to the 2,6-DIPN and 2,7-
DIPN.
Results of IPN disproportionation over the HB catalyst
(Table 5) are similar but clearer than those obtained on the
HY zeolite and also indicate shape-selectivity effects. The
concentration of bulky 1,3-DIPN, 1,6-DIPN, 1,7-DIPN, and
other isomers in the product obtained over HB zeolite was
even smaller than that obtained over HM-1 catalyst. 2,7-
DIPN and 2,6-DIPN were the most abundant isomers in the
DIPN product; however, the 2,6-DIPN/2,7-DIPN ratio was
still lower than that observed on HY zeolites and a ratio as
low as 0.34 was observed.
Such a small 2,6-DIPN/2,7-DIPN ratio indicates un-
doubtedly the shape-selectivity effect over HY and HB zeo-
lites. It is not observed over amorphous aluminosilicate and
it is out of the range predicted by thermodynamics [3,4] and
kinetics [4,11,15] of the reactions occurring in the tested sys-
tem.
Low concentrations of bulky DIPN isomers and high con-
centrations of slim 2,6-DIPN and 2,7-DIPN in the products
obtained over all tested zeolites can easily be explained by
steric hindrances in the pores. However, a question arises
why over HB and HY zeolites high selectivity toward 2,7-
DIPN is observed but over mordenite catalysts high 2,6-
DIPN selectivity occurs. Different acidity of the catalysts
seems not to be involved in this difference in selectivity,
because 2,6-selectivity was observed both on low and on
high-silica mordenites and 2,7-selectivity was observed on
low-silica HY and high-silica HB zeolites.
Fig. 6. Scheme of space-fill models of transition-state complexes leading to
2,7-DIPN (1 and 2) and 2,6-DIPN (2 and 3) placed in a cross section of Y
zeolite lattice.
available void space in the zeolite pores; therefore, this ex-
planation can rationalize the difference for favoring 2,6- or
2,7-DIPN. Schemes shown in Figs. 6 and 7 visualize space-
fill models of intermediate complexes leading to 2,6-DIPN
and 2,7-DIPN in pores of zeolite Y and mordenite, respec-
tively. Zeolite pores are shown as cross sections of the ze-
olite lattices. There are possibly several conformers of each
transition-state complex; however, the most distinguishing
examples of the complexes were chosen to visualize our
reasoning. Other conformers (e.g., having mutually perpen-
dicular or crossed planes of naphthalene rings) also seem to
obey this way of explanation.
The crystalline lattice of Y zeolite creates large void α-
cages ca. 1.1 nm in diameter, connected by 12-membered
windows of 0.74 nm in diameter [23]. The complex mole-
cules leading to 2,7-DIPN and 2,6-DIPN are ca. 1.8–1.9 and
2 nm long, respectively. There is not enough space in one
α-cage of Y zeolite to hold such long molecules. They must
occupy partially also the neighbor α-cage through the con-
necting window. It is evident from Fig. 6 that the molecules
of complexes 1 and 2, leading to 2,7-DIPN, are bent and bet-
ter fit the void space of the Y zeolite than linear molecules
of complexes 3 and 4, leading to 2,6-DIPN. It seems that
complexes 1 and 2 can be better arranged in the tight win-
dow connecting two α-cages of Y zeolite than complexes 3
and 4.
We carefully considered disproportionation mechanisms
and sizes of molecules and transition-state complexes and
we are able to propose an explanation as follows.
There are two possible mechanisms of isopropylnaph-
thalene disproportionation. First, the mechanism may be
monomolecular, i.e., a dealkylation–realkylation mecha-
nism. Isopropyl substituent (carbocation) abstracted from
the IPN molecule alkylates other molecules of IPN to form
DIPN. In spatially hindered pores of the tested zeolites the
2-IPN is dealkylated to naphthalene and the abstracted iso-
propyl substituent attacks the neighbor molecule of 2-IPN
at positions 6 or 7 to give respective DIPN isomers. The
2,6-DIPN/2,7-DIPN ratio in the product is expected to be
similar to that of alkylation of 2-IPN with propylene. The lit-
erature data on naphthalene isopropylation over HY and H-
beta zeolites [6,7,16,19,22] indicate that the 2,6-DIPN/2,7-
DIPN ratio is higher or close to 1. Also in our experiments
on naphthalene alkylation, performed over the same samples
of zeolite Y and beta in the same temperature range as dis-
proportionation, a ratio higher or close to 1 was observed.
Therefore, high 2,7-DIPN selectivity over HB and HY cata-
lysts cannot be explained by a monomolecular mechanism.
The second possibility is a bimolecular mechanism of
disproportionation. Two neighbor molecules of 2-IPN form
one huge molecule of intermediate complex by sharing one
of the isopropyl groups. Such complex molecules are long
and their shape can be strongly influenced by the shape of
The situation in the mordenite channels is different
(Fig. 7) and complexes leading to 2,6-DIPN (3 and 4) better
suit the shape of straight channels than complexes leading to
2,7-DIPN (1 and 2).
In other words, the shape of the complex being formed
from two IPN molecules depends on the shape of space

20
R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
positioned in the channel, along its axis with the bent part of
the molecule nested in the pocket.
Composition of the product obtained in the IPN dispro-
portionation is a consequence of many simultaneous reac-
tions and phenomena. The effect of the pore shape on the
2,6-DIPN/2,7-DIPN ratio can be concealed by the non-
shape-selective reactions which occur on the external surface
of the catalyst as well as by the reactions occurring inside the
pores. Such influence is well visible when reaction tempera-
ture is changed.
The dominating mechanism of disproportionation varies
with temperature. At low temperatures the bimolecular
mechanism of disproportionation dominates whereas at high
temperatures the monomolecular one dominates [27]. There-
fore, at low temperatures, the 2,6-DIPN/2,7-DIPN ratio is
strongly influenced by the shape of zeolite pores whereas at
300 ◦C the 2,6-DIPN/2,7-DIPN ratio approaches thermody-
namic equilibrium. A relatively high content of propylene
and aliphatics in the products of IPN disproportionation
carried out at 300 ◦C seems to confirm the monomolecular
(dealkylation/realkylation) mechanism.
Considering the bimolecular mechanism of dispropor-
tionation some comments are necessary to the results of
quantum chemical calculations and computer simulations
presented by Tasi et al. [4] on page 6516 and linked to
Ref. [17]. They stated that intermolecular isomerization of
DIPN in the pores of mordenite was impossible because
in the transition state of an intermolecular transfer the mi-
grating alkyl group was located between two naphthalene
rings and the transition complex was very large [4,17]. This
conclusion is in contradiction with our results of IPN dispro-
portionationand also with results obtained in other reactions,
e.g., disproportionation of cumene [28] or n-propylbenzene
[27] on zeolites, where a bimolecular mechanism was ob-
served. According to the Fig. 7 in Ref. [17], in the case of
methyl group transfer between two naphthalene rings the
space required for the reaction to proceed is greater than
1.3 nm (assuming 0.15 nm van der Waals radius for carbon
atom) and greater than the size of aromatic ring in the plane.
In the case of methyl group transfer between two naphtha-
lene rings a space of at least 0.71×0.88×1.3 nm is required
which is larger than any void space available in the mor-
denite pores. Similarly, the space necessary for bimolecular
reactions of benzene alkyl derivatives would be very large.
We must remember that calculations presented by Tasi
and co-workers [4,17] simulate two separate alkylnaph-
thalene molecules reacting in an unhindered environment
(without catalyst, solvent, or surrounding molecules) and
the arrangement with the lowest possible energy was deter-
mined. However, in narrow mordenite channels, while alkyl-
naphthalene molecules are approaching each other, defor-
mations of bonds connecting protons and alkyl substituents
with aromatic ring are possible. In most cases even in an
unhindered environment deformations of bonds of reacting
molecules can be energetically less demanding than mov-
ing bulky molecules to the position with the lowest en-
Fig. 7. Scheme of space-fill models of transition-state complexes leading
to 2,7-DIPN (1 and 2) and 2,6-DIPN (3 and 4) placed in a cross section of
mordenite channels.
available in the pores of zeolites. According to suggestions
by Derouane and co-workers [8,25] there are two possibili-
ties that force the complex shape and they both seem to lead
to diversity in 2,6-DIPN/2,7-DIPN ratios in pores of differ-
ent zeolites: first, the constraints in the arrangement of the
complex in the pores, and second, based on attractive inter-
actions (responsible for the confinement effect), i.e., match-
ing the shape of the complex to the shape of the pore walls.
In HY zeolite pores the complex would be then “wrapped
around” the hexagonal prism and sodalite cage, whereas in
“straight pipe”-type channels of mordenite the linear com-
plex is preferred.
Although the structure of beta zeolite is frequently not
uniform [23] the reason for high 2,7-DIPN selectivity is sim-
ilar to that described for zeolite Y; i.e., the more bent shape
of the transition-state complex better matches the void space
in the pores. Apart from straight channels the beta zeolite has
perpendicular, intersecting channels which are sinusoidal in
shape [24,26]. The complex leading to 2,7-DIPN can be
better arranged inside these channels or in the channel in-
tersections than the 2,6-DIPN precursors.
It is also possible that IPN disproportionation (or other
bimolecular reactions) influenced 2,7-DIPN selectivity ob-
served by Horsley et al. [8] in the alkylation of naphthalene
with propylene over L zeolite. In place of the tilted 2,7-DIPN
molecule nested in the side pocket of the L zeolite channel
shown in Fig. 5 in Ref. [8], a bent bimolecular complex lead-
ing to 2,7-DIPN can be easily located. The complex can be

R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
21
from shape selectivity in alkylation as well as from shape
selectivity in IPN disproportionation or DIPN isomerization
which occurs simultaneously with alkylation. For example,
two 2-IPN molecules, formed in the pores by naphthalene
alkylation, can disproportionate to naphthalene and β, β-
DIPN. The 2,7-DIPN molecule, which diffuses more slowly
than the 2,6-DIPN molecule [4,5,8], can undergo intramole-
cular isomerization to 2,6-DIPN [4] and intermolecular iso-
merization, i.e., dealkylation/realkylation and disproportion-
ation/transalkylation reactions with neighbor molecules of
naphthalene or IPN present in the pores of mordenite. For
example, the 2,7-DIPN molecule can disproportionate with
neighbor 2-IPN molecule to form 2,6-DIPN or another 2,7-
DIPN molecule and 2-IPN (formed from the starting 2,7-
DIPN). Such reactions of alkyl substituent transfer between
aromatic nuclei can be repeated several times during the mi-
gration of this group of molecules through the zeolite pores.
The 2,7-DIPN molecule can also transalkylate with neigh-
bor naphthalene molecule to form two molecules of 2-IPN
which in turn can disproportionate back to DIPN and naph-
thalene. If, under given conditions, a bimolecular mecha-
nism of disproportionation prevails over a monomolecular
mechanism the product obtained inside mordenite pores is
enriched in the 2,6-DIPN whereas those obtained over Y and
beta zeolites are enriched in the 2,7-DIPN. It is, therefore,
possible that 2,6-DIPN selectivity in alkylation is amplified
by the consecutive bimolecular reactions in mordenite pores
but is suppressed in pores of other wide pore zeolites.
Fig. 8. Scheme of arrangement of neighbor pairs of molecules of β,β-DIPN
(A) and 2-IPN (B and C) in mordenite channels. Bimolecular reaction in
case A and C is impossible whereas in case B it is feasible.
ergy. Moreover, at high temperatures not only deformations
of bonds are common but also bond cleavage is possible.
Therefore, in our opinion, the optimization of arrangement
of reacting alkylnaphthalene molecules to achieve the low-
est energy of the system is not evidence for the lack of a
bimolecular mechanism of DIPN isomerization and IPN dis-
proportionation.
An alternative explanation of the possible difficulties in
proceeding bimolecular isomerization of DIPNs is shown
schematically in Fig. 8. Two molecules of DIPN, reacting
in a narrow mordenite channel, have two neighboring alkyl
groups which hinder each other to attack a naphthalene ring
(Fig. 8A). In Fig. 8B 2-IPN disproportionation is unhindered
and the alkyl substituent can be easily transferred between
two neighbor naphthalene rings. However, a bimolecular dis-
proportionation is difficult or impossible in mordenite pores
if two IPN molecules are arranged to have two alkyl sub-
stituents neighboring (Fig. 8C) or no alkyl group between
both naphthalene nuclei. In order to enable the bimolecu-
lar reaction of DIPN (isomerization, disproportionation, or
transalkylation) in the zeolite pores, the presence of the
molecule with an easily accessible position in the aromatic
ring (naphthalene or IPN) in the surroundings of the DIPN
molecule is necessary.
4. Conclusions
In tests of disproportionation of isopropylnaphthalene
carried out over wide pore zeolites high β, β-selectivity
in diisopropylnaphthalene product was observed. However,
the 2,6-DIPN/2,7-DIPN mole ratio was dependent on the
structure of the applied zeolite and on the temperature.
Over H-mordenites 2,6-DIPN was the most preferred iso-
mer, whereas 2,7-DIPN was favored over HY and H-beta
zeolites.
Such differences in isomer preference can be explained
by a bimolecular mechanism of disproportionation. Depend-
ing on the zeolite pore architecture the bent transition-state
complex, leading to 2,7-DIPN or more linear, leading to 2,6-
DIPN, is formed.
Results of our investigations indicate that IPN dispropor-
tionation and other bimolecular reactions can contribute to
shape-selectivity effects observed during naphthalene alky-
lation carried out over zeolites. Relatively high activity of
zeolite catalysts in IPN disproportionation and compositions
of disproportionation products obtained even at a very short
contact time (viz. Tables 1–5) indicate how strongly the
product of naphthalene alkylation can be influenced by iso-
merization and disproportionation.
Reaction temperature determines which mechanism of
disproportionation, monomolecular or bimolecular, is dom-
inate. At low temperatures the bimolecular mechanism of
disproportionation is favored whereas at high temperatures
the monomolecular one (dealkylation/realkylation) is domi-
nate. In the latter case the 2,6-DIPN/2,7-DIPN mole ratio in
the product approaches an equilibrium value.
Relatively high activity of the tested catalysts in IPN
disproportionation indicates that secondary reactions can
strongly influence the shape-selectivity effect observed in
alkylation of naphthalene. In reactions carried out over ze-
High 2,6-DIPN/2,7-DIPN ratios observed in the alky-
lation tests carried out over mordenite catalysts can result

22
R. Brzozowski, W. Skupin´ski / Journal of Catalysis 220 (2003) 13–22
olite catalysts the pore size is not always the most critical
factor but the shape of the pores (curvatures, cavities, and
pockets) can also play an important role in shape selectivity.
[9] J.-H. Kim, Y. Sugi, T. Matsuzaki, T. Hanaoka, Y. Kubota, X. Tu, M.
Matsumoto, Micropor. Mater. 5 (1995) 113.
[10] J.-H. Kim, Y. Sugi, T. Matsuzaki, T. Hanaoka, Y. Kubota, X. Tu, M.
Matsumoto, S. Nakata, A. Kato, G. Seo, C. Pak, Appl. Catal. A 131
(1995) 15.
[11] Ch. Song, C. R. Acad. Sci. Paris, Ser. IIc 3 (2000) 477.
[12] A.D. Schmitz, Ch. Song, Catal. Today 31 (1996) 19.
[13] A.D. Schmitz, Ch. Song, Prepr. Pap., Am. Chem. Soc., Div. Fuel
Chem. 40 (1995) 918.
[14] A.D. Schmitz, Ch. Song, Catal. Lett. 40 (1996) 59.
[15] Ch. Song, X. Ma, A.D. Schmitz, H.H. Schobert, Appl. Catal. A 182
(1999) 175.
[16] P. Moreau, A. Finielis, P. Geneste, J. Joffre, F. Moreau, J. Solofo, Catal.
Today 31 (1996) 11.
[17] G. Tasi, F. Mizukami, M. Toba, S. Niwa, I. Palinko, J. Phys. Chem.
A 104 (2000) 1337.
Acknowledgments
This work was supported by the State Committee of Sci-
entific Research in Poland (KBN) by Grant 3 T09B 076 14
and by a grant for statutory activity of ICRI. We thank Süd-
Chemie GmbH (Germany) for providing the catalyst sam-
ples. We are also very grateful to the Journal of Catalysis
reviewers for their important corrections and discussions.
[18] R. Brzozowski, W. Skupin´ski, J. Catal. 210 (2002) 313.
[19] S.-J. Chu, Y.-W. Chen, Ind. Eng. Chem. Res. 33 (1994) 3112.
[20] I.-M. Tseng, J.-F. Wu, Y.-W. Chen, React. Kinet. Catal. Lett. 63 (1998)
359.
[21] R. Brzozowski, W. Skupin´ski, M.H. Jamróz, M. Skarz˙yn´ski, H. Otwi-
nowska, J. Chromatogr. A 946 (2002) 221.
[22] P. Moreau, A. Finielis, P. Geneste, J. Solofo, J. Catal. 136 (1992) 487.
[23] J.B. Nagy, P. Bodart, I. Hannus, I. Kiricsi, Synthesis, Characterization
and Use of Zeolitic Microporous Materials, DecaGen, Szeged, Hun-
gary, 1998.
[24] International Zeolite Association, Atlas of Zeolite Framework Types,
5th revised ed., 2001; www.iza-structure.org/databases.
[25] E.G. Derouane, J. Mol. Catal. A 134 (1998) 29.
[26] J.B. Higgins, Catal. Today 19 (1994) 7.
References
[1] D.E. Stuetz, European patent 172012, 1991.
[2] D. Mravec, M. Hronec, M. Michvocik, P. Moreau, A. Finielis, P. Gen-
este, Petrol. Coal 37 (1996) 63.
[3] R. Brzozowski, J.Cz. Dobrowolski, M.H. Jamróz, W. Skupin´ski,
J. Mol. Catal. A 170 (2001) 95.
[4] G. Tasi, F. Mizukami, I. Palinko, M. Toba, A. Kukovecz, J. Phys.
Chem. A 105 (2001) 6513.
[5] E. Kikuchi, K. Sawada, M. Maeda, T. Matsuda, Stud. Surf. Sci.
Catal. 90 (1994) 391.
[6] A. Katayama, M. Toba, G. Takeuchi, F. Mizukami, S. Niwa, S. Mita-
mura, J. Chem. Soc., Chem. Commun. 1991 (1991) 39.
[7] I. Sugi, M. Toba, Catal. Today 19 (1994) 187.
[8] J.A. Horsley, J.D. Fellmann, E.G. Derouane, C.M. Freeman, J. Ca-
tal. 147 (1994) 231.
[27] T.-C. Tsai, I. Wang, J. Catal. 133 (1992) 136.
[28] T.-C. Tsai, C.-L. Ay, I. Wang, Appl. Catal. 77 (1991) 199.