Acidity and catalytic properties of realuminated zeolite Y

Article abstract of DOI:10.1021/jp970900a

We have compared the acidity and the catalytic properties of (i) zeolite H-Y, (ii) H-Y dealuminated by hydrothermal treatment (H-US-Y), and (iii) H-US-Y realuminated by the KOH treatment (H-Real-US-Y). Quantitative IR measurements of pyridine sorption show that the concentration of "3460 cm^-1 hydroxyl groups", the Br?nsted acid sites active in many catalytic reactions, is slightly lower in sample H-Real-US-Y than in H-Y. This demonstrates that most of the 3460 cm^-1 hydroxyl groups, present in the parent material and subsequently destroyed by dealumination, are restored by realumination and the K⁺/H⁺ exchange. The turnover frequency of the realuminated zeolite in m-xylene reactions is higher than in the parent sample, as a result of the higher strength of the acid sites, caused by the higher population of the most strongly acidic Si₃Si-OH-AlSi₃ groupings. Dealumination and realumination also alter the selectivity in the transformation of m-xylene. Disproportionation of xylenes to toluene and trimethylbenzenes, which involves bulky transition step complexes, is hindered in the dealuminated zeolite, where the channels are partly blocked by the extra-framework A1 species, but is restored in the realuminated material, in which the extra-framework species are re-inserted into the framework.

Full text of DOI:10.1021/jp970900a

J. Phys. Chem. B 1997, 101, 6929-6932
6929
Acidity and Catalytic Properties of Realuminated Zeolite Y
Bogdan Sulikowski,*,† Jerzy Datka,‡ Barbara Gil,‡ Juliusz Ptaszynski,† and Jacek Klinowski*,§
Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, Niezapominajek 1, 30-239 Krako´w,
Poland, Faculty of Chemistry, Jagiellonian UniVersity, Ingardena 3, 30-060 Krako´w, Poland, and Department
of Chemistry, UniVersity of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K.
ReceiVed: March 11, 1997X
We have compared the acidity and the catalytic properties of (i) zeolite H-Y, (ii) H-Y dealuminated by
hydrothermal treatment (H-US-Y), and (iii) H-US-Y realuminated by the KOH treatment (H-Real-US-Y).
Quantitative IR measurements of pyridine sorption show that the concentration of “3460 cm-1 hydroxyl groups”,
the Brønsted acid sites active in many catalytic reactions, is slightly lower in sample H-Real-US-Y than in
H-Y. This demonstrates that most of the 3460 cm-1 hydroxyl groups, present in the parent material and
subsequently destroyed by dealumination, are restored by realumination and the K+/H+ exchange. The turnover
frequency of the realuminated zeolite in m-xylene reactions is higher than in the parent sample, as a result of
the higher strength of the acid sites, caused by the higher population of the most strongly acidic Si3Si-OH-
AlSi3 groupings. Dealumination and realumination also alter the selectivity in the transformation of m-xylene.
Disproportionation of xylenes to toluene and trimethylbenzenes, which involves bulky transition step complexes,
is hindered in the dealuminated zeolite, where the channels are partly blocked by the extra-framework Al
species, but is restored in the realuminated material, in which the extra-framework species are re-inserted
into the framework.
Introduction
+ [Al3+]EF, where the subscripts F and EF denote framework
and extra-framework atoms, respectively. The forward reaction
The Brønsted acidity of zeolites arises from the presence in
their hydrogen forms (usually prepared by the calcination of
the ammonium forms) of Si-OH-Al groups (“bridging hy-
droxyl groups”). Much attention has therefore been devoted
to the determination of their number and structure. It is clear
that the framework aluminum content must affect the population
and the strength of these acid sites, responsible for the catalytic
activity in many important reactions. The Si/Al ratio can be
controlled either during crystallization or via various postsyn-
thesis treatments. Zeolites X and Y (synthetic equivalents of
the rare mineral faujasite) can be synthesized with a very limited
range of Si/Al ratios (ca. 1.0 to ca. 3.4).1 The use of crown
ethers in the course of synthesis increases the upper limit to ca.
4.9.2 A number of methods aimed at increasing the Si/Al ratio
of zeolite Y have been developed.3,4 However, the most
common procedure is hydrothermal treatment at elevated
temperatures. Calcination of the ammonium form of zeolite
Y, either under deep bed conditions (self-steaming) or under
controlled atmosphere (steam, ammonia, etc.) yields a more
siliceous framework. Solid-state NMR monitors the chemical
status of aluminum expelled into the extra-framework positions5
during such treatment (known as “ultrastabilization”) and shows
that the resulting framework vacancies are reoccupied by silicon.
The presence of steam facilities the process. Materials prepared
in this way have valuable catalytic properties, and the majority
of fluid catalytic cracking catalysts (FCC) are manufactured by
steam treatment of zeolite Y.6
(i.e., dealumination) has been studied extensively.3,4 By
contrast, the reverse process, whereby the extra-framework
aluminum is re-inserted into the tetrahedral framework, has been
described only relatively recently in a handful papers.7-15 It
has been shown that treatment of dealuminated zeolite Y with
dilute aqueous solutions of bases (preferably KOH), yields
samples with an enhanced content of framework aluminum. 29
-
Si and 27Al magic-angle spinning (MAS) NMR, IR, and X-ray
and neutron diffraction show that, depending on the conditions,
most of the extra-framework aluminum can be re-inserted.7,8,13,15
Infrared spectra of hydrogen forms of as-prepared and
dealuminated zeolite Y (Si/Al > 2.00) contain two distinct
bands: from Si-O1H-Al hydroxyls (3631-3647 cm-1) and
Si-O3H-Al hydroxyls (ca. 3550 cm-1), where the subscripts
refer to the various crystallographically inequivalent kinds of
oxygen atoms in the faujasite framework. Although the precise
IR frequency of the former band depends somewhat on the Si/
Al ratio, we shall refer to it as the “3640 cm-1 band”. Our
recent IR study revealed16 that the same two bands appear in
the spectra of realuminated zeolite Y. This indicates that the
bridging hydroxyl groups destroyed by the hydrothermal treat-
ment are restored upon realumination with KOH followed by
the K+fH+ ion exchange. Most interestingly, benzene sorption
shows that the acid strength of the 3640 cm-1 OH groups in
realuminated H-Y is higher than before dealumination. We have
explained this in terms of the presence of different kinds of
bridging hydroxyl groups with respect to the number of
aluminum atoms connected, via oxygens, to the silicon involved
in an acidic hydroxyl. There are thus Si3Si-OH-AlSi3,
AlSi2Si-OH-AlSi3, and Al2SiSi-OH-AlSi3 hydroxyls. In
the “Si(nAl) notation” used in the NMR spectroscopy of zeolites
(n e 4 is the number of aluminum connected to a silicon) these
correspond to Si(1Al), Si(2Al), and Si(3Al) environments,
respectively. 29Si MAS NMR reveals that the contribution of
Si(1Al) species increases as a result of the dealumination-
realumination cycle, which means that the population of the
The dealumination/realumination process can be regarded as
the isomorphous substitution of the type [Si4+, Al3+]F T [Si4+
]
F
* Address correspondence to these authors. Jacek Klinowski: telephone
+44-1223-33 65 14; FAX +44-1223-33 63 62; E-mail jk18@cam.ac.uk.
Bogdan Sulikowski: telephone +48-12-25 28 41; FAX +48-12-25 19 23;
e-mail ncsuliko@cyf-kr.edu.pl.
† Polish Academy of Sciences.
‡ Jagiellonian University.
§ University of Cambridge.
X Abstract published in AdVance ACS Abstracts, July 15, 1997.
S1089-5647(97)00900-0 CCC: $14.00 © 1997 American Chemical Society

6930 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Sulikowski et al.
TABLE 1: Concentrations of 3640 cm-1 Hydroxyl Groups
(in µmol/g) in the Parent Sample H-Y and the Realuminated
Sample H-Real-US-Y As Determined by Pyridine Sorption,
and the Turnover Frequency Number (TOF) in the
Transformation of m-Xylenea
most acidic Si3Si-OH-AlSi3 protons and the average acid
strength of the 3640 cm-1 hydroxyl groups, have increased.
The aim of this work is to examine the acidity and catalytic
activity of bridging hydroxyl groups. It is particularly important
to know whether all the 3640 cm-1 hydroxyl groups in zeolite
H-Y prior to dealumination could be restored upon realumination
and to clarify the effect of the dealumination-realumination
cycle on the catalytic properties. We have therefore determined
the concentration of 3640 cm-1 hydroxyl groups in zeolite Y
by quantitative IR studies of pyridine sorption.
concentration of 3640 cm-1
hydroxyl groups (µmol/g)
sample
TOF (s-1
)
H-Y (parent)
H-Real-US-Y
1290
950
0.064
0.081
a We estimate the experimental error at less than (5% of each value.
Catalytic activity of the samples was tested in isomerization
and disproportionation of m-xylene, an industrially important
process used as a test reaction for shape-selective catalysts.20-24
Two important reactions involving xylene isomers are catalyzed
by acid sites: isomerization (o-xylene T m-xylene T p-xylene)
and disproportionation (2 xylene T toluene + trimethylben-
zenes). The latter “transalkylation reaction” frequently occurs
on acid catalysts. Isomerization may proceed according to a
unimolecular (by consecutive 1,2-methyl shifts) or a bimolecular
mechanism.25,26 Bimolecular disproportionation of alkylaro-
matic molecules is mostly governed by the channel structure27
and proceeds much more easily in large-pore zeolites such as
zeolites Y, L, and beta, where the bulky transition-state
complexes can be accommodated, than in zeolites with narrow
pores, such as ZSM-5, ZSM-35, or NU-10. The selectivity for
xylene transformation reflects therefore the intrinsic properties
of the catalyst. In addition to isomerization and disproportion-
ation, a third pathway involves dealkylation of alkylaromatics,
producing toluene, benzene, and light paraffins.
We have followed the effect of the dealumination and
realumination on the catalytic behavior of zeolite H-Y in the
reactions of m-xylene. The catalytic activity of 3640 cm-1
groups in the parent, dealuminated, and realuminated samples
has been monitored in terms of the relative values of turnover
numbers. The effect of dealumination and realumination on
the intrinsic properties of the zeolite was monitored by measur-
ing the selectivity in the transformation of m-xylene.
Infrared Spectroscopy. For IR studies, the zeolites were
pressed into thin wafers (3-8 mg/cm2 and activated in situ in
the IR cell at 720 K in vacuum for 1 h. The spectra were
recorded with a Bruker 48 PC spectrometer equipped with an
MCT detector.
Catalytic Tests. Zeolite crystals were pressed into binder-
free wafers, crushed and sieved to 30-60 mesh. A 50 mg
sample was diluted with 0.5 mL inactive glass spheres and
packed into a stainless steel down-flow microreactor (10 mm
i.d.) placed in a tubular furnace. Additional amounts of glass
spheres (0.5 mL) were placed on top of the catalyst bed. The
catalyst was heated to 723 K at a rate of 100 K/h and activated
in a nitrogen flow of 50 mL/min for 3 h. Measurements were
performed in the pulse mode, in order to avoid coke formation
which can obscure the intrinsic activity and selectivity of the
catalysts. All reaction parameters were kept constant during
the test to maintain maximum reproducibility. The products
were analyzed with a GCHF gas chromatograph equipped with
a TCD detector and a 3 m column packed with Bentone-34,
didecyl phthalate, and silicon oil A on Chromosorb W (30-60
mesh). We estimate the error involved in product analysis at
less than 0.1 mol%.
Results and Discussion
Brønsted Acidity. We have monitored the concentration of
3640 cm-1 hydroxyl groups (projecting into supercage and thus
accessible to the reacting molecules) by measuring the sorption
of pyridine. In the parent sample H-Y this gives rise to a band
from pyridinium ions PyH+ (1545 cm-1) and decreases the
intensity of the 3640 cm-1 band. Small portions of pyridine
were adsorbed quantitatively at 420 K, and the extinction
coefficient was then calculated from the linear dependence of
intensity of the 1545 cm-1 band versus the amount of pyridine
sorbed. The resulting value of 0.085 cm2 µmol was used to
determine the concentration of pyridinium ions in samples in
which all the high-frequency OH groups are neutralized by
pyridine. The results were 1290 µmol/g for the parent sample
H-Y and 950 µmol/g for the realuminated sample H-Real-US-Y
(Table 1). We estimate the experimental error at less than (5%
of each value. The decrease in concentration after the comple-
tion of the dealumination-realumination cycle is caused by not
all aluminum atoms reentering the framework.
Catalytic Activity. The parent sample H-Y is highly active
in the transformation of m-xylene (Table 2). Toluene and
trimethylbenzenes (TMB) are formed in addition to the products
of isomerization. Toluene is produced by disproportionation
and dealkylation of xylene isomers. The contribution of these
two parallel reactions, with higher energy of activation than that
required for isomerization, increases with temperature. In
sample H-Y the concentration ratio of the products of isomer-
ization (p-xylene + o-xylene) to that of disproportionation
(toluene + TMB) decreases with temperature from 3.85 to 673
K to 1.50 to 723 K. The p-xylene/o-xylene concentration ratio
remains essentially constant (0.64-0.68) in this temperature
Experimental Procedures
Catalyst Preparation. The ultrastable sample H-Y was
prepared by hydrothermal treatment17,18 of highly crystalline
synthetic zeolite Na-Y with Si/Al ) 2.47 and Na/Al ) 1.09 (as
determined by atomic absorption). The 78% ammonium-
exchanged form, prepared by stirring Na-Y twice with 10 wt%
NH4Cl at 368 K for 1 h, had the lattice parameter ao ) 2.466
nm. The ammonium form was heated in a vertical quartz reactor
at 823 K for 18 h, with water being injected at a rate of 12
mL/h so that the partial pressure of H2O above the sample was
1 atm. The hydrothermally dealuminated form H-US-Y had
the lattice parameter ao ) 2.451 nm.
A 2 g portion of H-US-Y was realuminated by stirring with
100 mL of 0.25 M aqueous solution of KOH at 353 K for 25
h. The product was washed with water and dried. KOH rather
than NaOH was used because synthetic faujasites cannot
recrystallize from potassium-bearing solutions.19 Since after
realumination the sample was in the potassium form, it was
exchanged four times with ammonium nitrate to yield NH4-
Real-US-Y (ao ) 2.463 nm). Thermal treatment of this sample
gave H-Real-US-Y, suitable for catalytic testing.
X-ray Diffraction. Powder X-ray diffraction patterns were
acquired on Dron-2 and Philips diffractometers using Ni-filtered
Cu KR radiation. All samples were hydrated over saturated
aqueous calcium nitrate prior to measurement. Silicon powder
was used as an internal standard for lattice parameter calcula-
tions.

Properties of Realuminated Zeolite Y
J. Phys. Chem. B, Vol. 101, No. 35, 1997 6931
TABLE 2: Transformation of m-Xylene on Parent Zeolite H-Y, Dealuminated (Ultrastable) Sample H-US-Y, and the
Hydrogen Form of Realuminated Sample (H-Real-US-Y)a
673 K
H-US-Y
698 K
H-US-Y
723 K
H-US-Y
product
benzene
toluene
ethylbenzene
p-xylene
m-xylene
H-Y
H-Real-US-Y
H-Y
H-Real-US-Y
H-Y
H-Real-US-Y
0.74
11.26
0.74
11.24
55.95
16.47
0.76
2.46
0.38
3.60
27.71
0.68
0.45
8.83
0.90
2.05
80.01
5.86
0.13
1.17
traces
1.30
7.91
0.35
3.04
4.42
24.15
0.11
6.35
51.50
8.86
1.22
3.39
traces
4.61
15.21
0.72
1.65
1.76
20.35
0.04
10.48
45.45
15.48
1.30
4.28
0.86
6.44
25.96
0.68
0.82
12.67
0.78
2.28
77.25
5.49
0.18
0.53
traces
0.71
7.77
0.42
5.47
5.73
25.55
0.11
6.41
48.32
9.22
1.59
3.07
traces
4.66
15.63
0.70
1.68
3.16
26.54
0.01
1.39
15.38
0.77
2.84
69.06
9.36
0.29
0.91
traces
1.20
12.20
0.30
5.08
7.56
28.84
0.33
6.09
43.27
8.19
1.21
4.51
traces
5.72
14.28
0.74
1.25
8.87
40.08
13.79
1.55
5.14
0.86
7.55
22.66
0.64
o-xylene
1,3,5-TMB
1,2,4-TMB
1,2,3-TMB
∑TMB
p-xylene + o-xylene
p-xylene/o-xylene
I/D
3.85
2.02
1.50
a Product distribution is given in mol%. TMB ) trimethylbenzene, I ) isomerization, D ) disproportionation, and I/D ) (p-xylene + o-xylene)/
2∑TMB. The number of toluene molecules formed via transalkylation is equal to the sum of trimethylbenzene isomers. We estimate the error
involved in product analysis at less than 0.1 mol %.
range. We note that the experimentally observed distribution
of trimethylbenzenes is close to that required by thermodynamic
equilibrium arguments.28 Finally, small amounts of benzene
(0.74-3.16%) are formed as a result of consecutive reactions:
dealkylation of toluene and transalkylation of toluene and xylene
isomers.
The conversion of m-xylene on the dealuminated sample
H-US-Y is much lower (20-30%) than on parent H-Y. This
is in agreement with the IR studies, which show a distinct
concentration decrease of the catalytically active 3640 cm-1
hydroxyl groups in the former sample.16 Dealkylation and
disproportionation are hindered, so that the contribution of
isomerization is higher. The isomerization/disproportionation
ratio (3.04-5.08) shows a smaller contribution of the dispro-
portionation route, demanding a bimolecular transition state. It
is therefore clear that the aluminum species in the extra-
framework positions decrease the void space available for the
reactants.
The overall conversion of m-xylene rises significantly upon
realumination and is essentially the same in H-Real-US-Y as
in H-Y (Table 2), despite the lower concentration of acid centres
available for the reactant on the realuminated catalyst (Table
1). We also note that the amounts of TMB in sample H-Real-
US-Y are close to that found for H-Y, while more toluene is
formed. The isomerization/disproportionation (I/D) ratio is only
1.25-1.65 (Table 2). In other words, the contribution of the
space-demanding bimolecular transalkylation of m-xylene in-
creases again after realumination, and the relative contributions
of the two reaction pathways become similar to that observed
for the parent H-Y. As extra-framework aluminum re-enters
the framework, the contribution of isomerization decreases at
the expense of transalkylation route. Finally, shape selectivity,
expressed in terms of the p-xylene/o-xylene ratio, remains the
same within the experimental error, and recalls that of the parent
sample H-Y. No steric restrictions operate therefore in the
realuminated sample, which supports the proposed mechanism
of realumination.7 The I/D ratio thus appears as a very useful
tool for the characterization of modified zeolite catalysts.
Having determined the concentration of the active acid sites
(3640 cm-1 hydroxyl groups), we can make a quantitative
comparison of the catalytic performance of the samples. The
turnover frequency (TOF) of m-xylene transformation calculated
taking into account the number of proton sites accessible for
reacting molecules, is 25% higher for the realuminated sample
(Table 1). This indicates a higher acid strength of the
catalytically active hydroxyl groups, in agreement with the
results of our study of benzene sorption.16 Finally, the TOF
values are close to the value of 0.074 s-1 measured by Corma
et al.29 for zeolite H-Y the Si/Al ) 4.3. We conclude that the
increased acid strength of bridging hydroxyl groups16 in the
realuminated material is indeed caused by the higher population
of the Si(1Al) groups (and thus of the Si3Si-OH-AlSi3
hydroxyls) in comparison with sample H-Y, as monitored by
29Si MAS NMR.7
Acknowledgment. This work was partly supported by the
grant from the Polish Komitet Badan Naukowych (Project No.
0634/P3/94/07).
References and Notes
(1) Kacirek, H.; Lechert, H. J. Phys. Chem. 1976, 80, 1291.
(2) Delprato, F.; Delmotte, L.; Guth, J. L.; Huve, L. Zeolites 1990,
10, 546.
(3) Scherzer, J. In Catalytic Materials: Relationship between Structure
and ReactiVity; Whyte, T. E., Jr., Dalla Betta, R. A., Derouane, E. G., Baker,
R. T. K., Eds.; ACS Symposium Series 248: American Chemical Society:
Washington, DC, 198; p 157.
(4) Sulikowski, B. Heterog. Chem. ReV. 1996, 3, 203 and references
therein.
(5) Klinowski, J.; Thomas, J. M.; Fyfe, C. A.; Gobi, G. C. Nature 1982,
296, 533.
(6) Rabo, J. A.; Pellet, R. J.; Risch, A. P.; Blackwell, C. S. Presented
at the Katalistiks’ 8th Annual Fluid Cat Cracking Symposium, June 1-4,
1987, Budapest, Hungary.
(7) Hamdan, H.; Sulikowski, B.; Klinowski, J. J. Phys. Chem. 1989,
93, 350.
(8) Sulikowski, B.; Rakoczy, J.; Hamdan, H.; Klinowski, J. J. Chem.
Soc., Chem. Commun. 1987, 1542.
(9) Chang, C. D.; Chu, C. T.-W.; Miale, J. N.; Bridger, R. F.; Calvert,
R. B. J. Am. Chem. Soc. 1984, 106, 8143.
(10) Anderson, M. W.; Klinowski, J.; Liu, X. J. Chem. Soc., Chem.
Commun. 1984, 1596.
(11) Liu, X.; Klinowski, J.; Thomas, J. M. J. Chem. Soc., Chem.
Commun. 1986, 582.
(12) Man, P. P.; Klinowski, J. J. Chem. Phys. Lett. 1988, 147, 581.
(13) Hamdan, H.; Klinowski, J. In Zeolite Synthesis; Occelli, M. L.,
Robson, H. E., Eds.; 196th National Meeting of the American Chemical
Society, Los Angeles, CA, September 25-30, 1988; ACS Symposium Series
398; American Chemical Society: Washington, DC, 1989; p 448.
(14) Hamdan, H.; Klinowski, J. ACS Symp. Ser. 1989, 398, 465.
(15) Barrie, P. J.; Gladden, L. F.; Klinowski, J. J. Chem. Soc., Chem.
Commun. 1991, 592.
(16) Datka, J.; Sulikowski, B.; Gil, B. J. Phys. Chem. 1996, 100, 11242.
(17) McDaniel, C. V.; Maher, P. K. In Molecular SieVes; Society of
Chemical Industry: London, 1968; p 186.
(18) Kerr, G. T.; Shipman, G. F. J. Phys. Chem. 1968, 72, 3071.
(19) Barrer, R. M. Hydrothermal Chemistry of Zeolites; Academic
Press: London, 1982.
(20) Sulikowski, B. React. Kinet. Catal. Lett. 1981, 16, 39.

6932 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Sulikowski et al.
(21) Mavrodinova, V.; Penchev, V.; Lohse, U.; Stach, H. Zeolites 1989,
9, 197.
(22) Dewing, J. J. Mol. Catal. 1984, 27, 25.
(23) Richter, M.; Fiebig, W.; Jerschkewitz, H. G.; Lischke, G.; O¨ hlmann,
G. Zeolites 1989, 9, 238.
(26) Corma, A.; Sastre, E. J. Chem. Soc., Chem. Commun. 1991, 594.
(27) Martens, J. A.; Perez-Pariente, J.; Sastre, E.; Corma, A.; Jacobs,
P. A. Appl. Catal. 1988, 45, 85.
(28) Taylor, W. J.; Wagman, D. D.; Williams, M. G.; Pitzer, K. S.;
Rossini, F. D. J. Res. Natl. Bur. Stand. 1946, 37, 95.
(24) Datka, J.; Piwowarska, Z.; Rakoczy, J.; Sulikowski, B. Zeolites
1988, 8, 199.
(25) Karge, H. G.; Ladebeck, J.; Sarbak, Z.; Hatada, K. Zeolites 1982,
2, 94.
(29) Corma, A.; Llopis, F.; Monton, J. B. In New Frontiers in Catalysis,
Proc. 10th Intern. Congress on Catalysis, Budapest, July 19-24, 1992;
Guczi, L., Solymosi, F., Te´te´nyi, P., Eds.; Akade´miai Kiado´: Budapest,
1993; Part B, p 1145.