You are on page 1of 53

Dealumination Techniques for Zeolites

Hermann K. Beyer
Institute of Chemistry, Chemical Research Center, Hungarian Academy of Sciences,
Pusztaszeri t 5967, 1025 Budapest, Hungary; e-mail: beyer@chemres.hu

Dedicated to Professor Gerhard Ertl on the occasion of his 65th birthday

1 Introduction and Scope . . . . . . . . . . . . . . . . . . . . . . . . . 204

2 Extraction of Framework Aluminum by Chemical Agents . . . . . . 205


2.1 Dealumination with Acids . . . . . . . . . . . . . . . . . . . . . . . . 205
2.2 Dealumination with Complexing Agents . . . . . . . . . . . . . . . . 208
2.3 Instability of Hydrogen Zeolites Towards Liquid Water . . . . . . . . 210
2.4 Gaseous Halogen Compounds as Dealuminating Agents . . . . . . . 211

3 Hydrothermal Dealumination of Zeolite Frameworks . . . . . . . . 213


3.1 Early Fundamental Investigations . . . . . . . . . . . . . . . . . . . 213
3.2 Review of Recent Investigations . . . . . . . . . . . . . . . . . . . . . 218
3.2.1 Faujasite-Type Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . 218
3.2.2 ZSM-5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
3.2.3 Other Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
3.3 Thermal Dealumination . . . . . . . . . . . . . . . . . . . . . . . . . 225

4 Isomorphous Substitution of Framework Silicon for Aluminum . . 226


4.1 Dealumination with Silicon Tetrachloride . . . . . . . . . . . . . . . 226
4.1.1 Faujasite-Type Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . 226
4.1.2 Other Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
4.2 Isomorphous Substitution with Other Silicon Halides . . . . . . . . 236
4.3 Dealumination with (NH4)2[SiF6] Solutions . . . . . . . . . . . . . . 237
4.4 Dealumination of Zeolites in Dry Mixtures with (NH4)2[SiF6] . . . . 240

5 Alumination of Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . 242


5.1 Alumination with Gaseous Aluminum Chloride . . . . . . . . . . . . 243
5.2 Alumination with Aqueous Fluoroaluminates . . . . . . . . . . . . . 244
5.3 Alumination with Aluminate Solutions . . . . . . . . . . . . . . . . . 244
5.4 Re-Insertion of Extra-Framework Aluminum . . . . . . . . . . . . . 245

6 Desilication of Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . 247

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248

Molecular Sieves, Vol. 3


Springer-Verlag Berlin Heidelberg 2002
204 H.K. Beyer

1
Introduction and Scope
The framework Si/Al atomic ratio of zeolites is an important parameter that
exerts a strong influence on properties such as maximum ion-exchange capaci-
ty, thermal and hydrothermal stability, hydrophobicity, concentration and
strength of acid sites of the Brnsted-type, which may be generated in zeolite
structures, and catalytic activity and selectivity. Though in some processes, e.g.,
in ion-exchange procedures, low Si/Al ratios may be preferred, generally zeolites
with low aluminum content (high Si/Al ratio) are more favorable, especially
when applied as catalysts. However, the framework Si/Al ratio of zeolites pre-
pared by direct synthesis, such as by direct hydrothermal crystallization in
absence of templates, is generally restricted to more or less narrow limits. One of
the technically most important members of the zeolite family, faujasite, cannot
be directly synthesized with Si/Al ratios substantially higher than 2.5, at least not
in economically reasonable crystallization times. It is, therefore, of great impor-
tance to find methods that can increase the Si/Al ratio by chemical post-synthe-
sis modification of zeolite frameworks, i.e., by dealumination.
In the strict sense of the word, the term dealumination refers to the removal
of aluminum from zeolite frameworks by chemical reactions generally resulting
in lattice deficiencies. However, in its general use, it relates to the more complex
process comprising the incorporation of other elements, especially of silicon,
into the transient framework vacancies left temporarily by the release of alu-
minum.
Processes which increase the Si/Al ratio of zeolite structures may be subdivi-
ded into three categories:
1. Those involving only the removal of framework aluminum by chemical
agents or in case of hydrogen forms of zeolites by thermal dehydroxylati-
on, thereby resulting in lattice defects.
2. Those including, in addition to the mere extraction of framework aluminum,
a second step in which framework vacancies are filled in, e.g., by intrinsic sili-
con and oxygen atoms migrating in the zeolite lattice under hydrothermal
conditions.
3. Those representing true substitution reactions between the aluminum com-
ponent of the framework and the dealumination agent, being a compound of
the element to be incorporated, such as silicon.
The present review is aimed at covering first the literature which has appeared
since about 1985 on both methodical and mechanistic aspects of dealumination
techniques and on their structural and compositional consequences; although
fundamental and pioneering contributions published before this date will also be
included. Progress made in the last decade in the field of alumination and desili-
cation of zeolites, i.e., of processes closely related in nature to the main subject,
will also be reviewed. However, this review does not refer to papers dealing exclu-
sively or predominantly with special properties (e.g., acidity, catalytic activity,
hydrophobicity) and applications of dealuminated zeolites and to the isomor-
phous replacement of framework aluminum by elements other than silicon.
Dealumination Techniques for Zeolites 205

The older literature on dealumination of zeolites has been extensively re-


viewed by Scherzer [1], Stach et al. [2] and, with emphasis on mordenite, by
Karge and Weitkamp [3]. Recently, Sulikowski [4] dealt with dealumination and
alumination of zeolites as part of a more general review.

2
Extraction of Framework Aluminum by Chemical Agents

2.1
Dealumination with Acids

Removal of aluminum from a zeolite framework was first reported in 1964 by


Barrer and Makki [5]. They extracted aluminum from clinoptilolite by refluxing
with hydrochloric acid. Depending on the acid concentration, up to 100% of the
framework aluminum could be removed; however, the thermostability of the
products gradually decreased at dealumination degrees higher than 65%. De-
alumination with acids was accompanied by ion exchange of lattice cations by
protons. The overall process was suggested to proceed according to Eq. (1) under
formation of defect sites later generally denoted by the term hydroxyl nest.

(1)

Dealumination with mineral acids was also successfully applied to erionite [6],
mordenite [7, 8], offretite [9] and ZSM-5 [10]. The dealumination of mordenite
with mineral acids was monitored by 27Al and 29Si MAS NMR spectroscopy [11]
and compared with other dealumination procedures (steaming, reaction with
SiCl4). At the beginning of the process, acid leaching generates, in agreement
with the stoichiometry of Eq. (1), about four SiOH groups per one Al extracted.
Further dealumination was found to lead to a reorganization of the structure,
even at 100C, as shown by the decreasing amount of defects. A hypothesis on
the location of framework aluminum in the mordenite structure and the deal-
umination mechanism was presented. Karge and Dondur [12] used ammonia
TPD to study the distribution of acidity in mordenites dealuminated by acid
leaching The influence of acid leaching with nitric acid of different concentra-
tions on unit cell parameters, relative crystallinity, adsorption behavior and
amount of framework and extra-framework aluminum species of mordenites
of different origin was investigated by van Niekerk et al. [13]. It was found that
the extent of dealumination and the amount of extra-framework aluminum
remaining in the zeolite channels was strongly influenced by the crystallite size
and that dealumination is associated with a partial loss in crystallinity. A sys-
tematic study of acid leaching of sodium and hydrogen-exchanged mordenite
has been reported [14] that deals with the formation and removal of extra-
framework aluminum species and the creation of acidity upon dealumination.
206 H.K. Beyer

The framework of beta zeolites was found to dealuminate upon mild acid
leaching procedures [15, 16]. Refluxing in 0.5 N and 1 N hydrochloric acid for 4 h
resulted in an increase of the bulk Si/Al ratio from 11.5 to 32 and from 19 to 70
for samples prepared in alkaline and fluoride medium, respectively [15]. The
extraction of aluminum from as-synthesized zeolite beta still containing the
tetraethylammonium template with nitric acid of different concentrations was
studied [16]. Samples of zeolite beta with Si/Al ratios greater than 1000 com-
prising mesopores and three different types of silanol defect groups were ob-
tained in a single step without significant loss of crystallinity, porous volume
or thermal stability.
Depending on the acid concentration, Nu-2 zeolite could be progressively
dealuminated up to a Si/Al ratio of 90 by treatment with hydrochloric acid at
room temperature [17]. The removal of aluminum was found to be associated
with a gradual but slight decrease in crystallinity (maximum 20%).
ZSM-5 was reported to release framework aluminum completely upon treat-
ment with 1 N hydrochloric acid [18]. In contrast, Kornatowski et al. [19] found
that ZSM-5 can be only partly dealuminated by acid treatment. Recently, Kooy-
man et al. [20] reported that the bulk aluminum content of ZSM-5 zeolites could
not be significantly decreased by extraction with 1 N hydrochloric acid even at
temperatures as high as 160C. From this it follows that the framework alumi-
num content was little affected since the amount of extra-framework aluminum
detected by 27Al MAS NMR spectroscopy after acid leaching at 80C, somewhat
dependent on the zeolite preparation mode, was found to be rather small. HBr
and H2SO4 proved to be even less effective than HCl. The high stability towards
dealumination by acid leaching is attributed to the virtual absence of structural
defects in the ZSM-5 samples studied.
It is obvious that dealumination of aluminum-rich zeolite frameworks result-
ing in the formation of high lattice defect concentrations should diminish the
stability of the crystal structure. As early as 1958 it was reported [21] that the
structure of faujasite-type zeolites collapsed completely upon treatment with
strong mineral acids. However, Lee and Rees [22] have shown that the crystal
structure of Y zeolite is not significantly affected if the amount of HCl applied in
aqueous solutions does not exceed 10 mmol/g Na-Y which results in the release
of 56% of the framework aluminum atoms and in the complete exchange of the
sodium cations. Thus, at least part of the aluminum in Na-Y zeolite can be
extracted without considerable lattice destruction if HCl is applied in amounts
that do not yet cause too intense dealumination.
For a considerable period of time, contradictory statements have been made
in the literature on the thermal stability of zeolites dealuminated by acid leach-
ing. It seems to be self-evident that the thermal dehydroxylation of hydroxyl
nests results in the formation of new Si-O-Si bonds and, hence, at least in local
rearrangements of the framework atoms in the neighborhood of the vacancies.
This reasoning is supported by the partial loss in crystallinity upon acid leach-
ing reported in many of the papers reviewed above. It is to be expected that
structures with higher concentrations of such defects are less thermostable than
undisturbed lattices. However, there are also several publications (cf. e.g., [8]) in
which it is explicitely stated that the extraction of framework aluminum resulted
Dealumination Techniques for Zeolites 207

100

2 1 3
Crystallinity (%)

50

300 700 1100 1300


Temperature (K)
Fig. 1. Relative intensities of the (150) XRD reflection of 1 parent Na-mordenite (Si/Al = 5.9),
2 after acid leaching (Si/Al = 35) and 3 after subsequent steaming at 600C for 2 h [23]

in a significant increase in the thermal stability. Most probably, in these studies,


the experiments were performed in such a way that the water steam formed
during the thermal treatment as a reaction product of the dehydroxylation of
hydroxyl nests remained in contact with the zeolite sample for a longer time
(e.g., in case of deep-bed calcination). Under these conditions the process of
ultrastabilization (see Sect. 3.1) may have occurred. This process probably
played the decisive role also in the formation of the secondary pore system with
channel diameters of 2 and 3.4 nm observed by Wolf and John [8] in mordenite
dealuminated by acid leaching.
For Na-mordenite it has definitely been shown [23] that the thermal stability
significantly decreased after removal of about 80% of the framework aluminum
by acid leaching but increased again and even surpassed considerably the stabi-
lity of the parent material after subsequent steaming. This is illustrated by the
dependence of the thus-called X-ray crystallinity, i.e., the intensities of select-
ed reflections, on the calcination temperature in Fig. 1. The intensity of the (150)
reflection, which proved to be nearly unaffected by acid leaching and steaming,
was selected as standard. Though the intensities of XRD reflections do not
depend solely on the crystallinity and, hence, do not always give correct infor-
mation on the lattice destruction, the relative crystallinity values plotted in Fig. 1
may be considered as informative since they are related to the respective starting
material and the chemical composition of the samples does not change basical-
ly during the calcination process.
208 H.K. Beyer

2.2
Dealumination with Complexing Agents

In 1968, Kerr prepared dealuminated Y zeolites by extraction of framework


aluminum from Na-Y with H4EDTA solution at ambient temperature [24]. Up to
50% of the aluminum could be removed without any substantial loss in crystal-
linity and it was claimed that the products showed improved thermal stability.
The dealumination mechanism proposed by Kerr [25] comprises:
1. hydrolysis of Si-O-Al bonds, obviously provoked by the acidity of the agent,
which results in the extraction of aluminum from the framework, and
2. solubilization and, hence, mobilization of the formed cationic non-frame-
work aluminum species by complexation with EDTA.
In line with this dealumination mechanism, no reaction was found to occur bet-
ween NaY zeolite and the non-acidic Na2H2EDTA.
Datka et al. [26] reported that leaching of Na-Y zeolite with a 0.4% solution
of H4EDTA at 100C for 1 h gives defective crystals with a framework Si/Al ratio
of about 3.2 in which framework vacancies created by the release of aluminum
are not healed by silicon from other parts of the crystals. Ciembroniewicz et al.
[27] presented evidence for the creation of a secondary pore system with pore
diameters of about 3 nm upon treatment of zeolite Y with H4EDTA, especially if
more than 40% of the framework aluminum was removed. It was suggested that
these mesopores were associated with a gel phase formed by gradual amor-
phization of zeolitic material during the process. Thus, this observed phenome-
non was not attributed to a secondary mesopore system inside the zeolite
crystals, as later found by Lohse et al. [28] in hydrothermally dealuminated Y
zeolite.
It seems that dealumination with chelating agents is essentially an acid-leach-
ing process where the effectiveness is enhanced by complexing of the aluminum
species formed as reaction products. The process proceeds stoichiometrically
(see Fig. 2) so that it can be controlled by the amount of H4EDTA calculated for
dealumination to the desired level [24, 29].
Other acidic chelating agents, e.g., acetylacetone [30, 31], tartaric acid [32]
and oxalic acid [3335], have been successfully applied for the dealumination of
zeolites.
According to a patent assigned to the Mobil Oil Corp. [36], up to 40% of the
aluminum content of zeolites with Si/Al ratios greater than 1.5 (zeolite Y, zeolite
T, erionite, clinoptilolite, phillipsite) can be removed without substantially de-
stroying crystallinity when the parent zeolite is subjected to reflux in an aqueous
solution of chromium salts, preferably of CrCl3. The pH of the solution should
be less than 3.5. Chromium was found to be incorporated into the zeolite partly
by ion exchange but also, e.g., into faujasite up to 6 wt.%, in a non-exchangeable
form. This procedure belongs to the family of processes dealt with in this section
since the dealuminating effect of chromium salt solutions is assumed to be due
to the solubilization of hydrolyzed aluminum species by formation of soluble
binuclear aquo-hydroxy complexes comprising both chromium and aluminum
as central atoms [37].
Dealumination Techniques for Zeolites 209

1.00

0.80
Fraction Al removed from zeolite

0.60

0.40

0.20

0
0 0.2 0.4 0.6 0.8 1.0
Moles H4EDTA used per F.W. NaY

Fig. 2. Stoichiometry of aluminum removal from Na-Y by H4/EDTA. FW formula weight,


NaAlO2(SiO2)y [24]

Garwood et al. [36, 38] reported that subsequent to dealumination with EDTA
a significant amount of silicon could be digested and removed by refluxing with
1 N NaCl (or other salt) solutions while silicon was less readily removable after
dealumination with chromium chloride. This behavior was considered to evi-
dence the incorporation of chromium into framework vacancies left after relea-
se of aluminum, resulting in healing of the lattice ruptures and, hence, in more
stable structures.
Liu and Xu [39] reported on a limited increase in the Si/Al ratio of NH4-Y
from about 2.5 to 3.4 upon treatment with 0.1 M aqueous solutions of NH4[BF4]
at 60C for 24 h. They suggested that aluminum is released from the framework
induced by slow hydrolysis of the boron complex and, subsequently, silicon is
incorporated into the lattice vacancies left by dealumination. The silicon in-
volved in the claimed healing process was believed to originate from dissolution
of atoms located on the external surface of the zeolite and from amorphous
silica (if any). However, in this case, the insertion of silicon into the framework
210 H.K. Beyer

is doubtful since the authors draw conclusions only from changes in the crystallo-
graphic lattice constant and the intensities of 29Si MAS NMR signals obtained
without applying 1H cross polarization. These methods alone do not allow defi-
nitive conclusions to be drawn concerning the Si/Al ratio after dealumination
processes, all the more so as a new IR band found at 3738 cm1 clearly indicated
the formation of silanol groups generally associated with lattice deficiencies. It
is highly probable that in the reported dealumination process aluminum is
extracted from the framework by acidity created by partial hydrolysis of the
boron complex and then removed from the zeolite due to conversion into the
soluble complex salt NH4[AlF4].
In conclusion, the complexing agent must be acidic in order to be applicable
for the removal of framework aluminum.

2.3
Instability of Hydrogen Zeolites Towards Liquid Water

H-Y zeolite, obtained by thermal deammoniation of the ammonium form, was


found to be thermally extremely unstable after resorption of water from the
atmosphere [40, 41]. This phenomenon was ascribed to a framework dealumina-
tion process similar to that observed upon acid leaching. Later, Beyer et al. [42]
observed that full rehydration of H,Na-Y (exchange degree 78%), obtained by
thermal deammoniation of the respective ammonium form, resulted in a con-
siderable loss in X-ray crystallinity upon heating at relatively low temperatures
(180C) and virtually affected the crystal structure already at ambient tempera-
ture inasmuch as the original ammonium form could not be fully re-obtained by
adsorption of ammonia. In contrast, partial re-adsorption of water (77 mg/g) did
not show any effect on the thermal stability of the lattice. Furthermore, the crystal
structure of hydrogen zeolites stable towards the attack of mineral acids (morde-
nite, clinoptilolite) proved to be not affected by readsorption of water. Thus, the
instability of H,Na-Y was attributed to the removal of framework aluminum or,
at least, hydrolytic cleavage of Si-O-Al bonds by the intrinsic acidity of the hydra-
ted H-zeolite which becomes manifest when lattice hydrogen atoms form H3O+
ions in intracrystalline liquid water. Confirming these findings, Karge [43] pro-
vided infrared spectroscopic evidence for the instability of the lattice of H,Na-Y
upon contact with water vapor. Maessen et al. [44] prepared H,Na-Y zeolites with
exchange degrees of 70 and 87% from the respective ammonium forms by low-
temperature plasma calcination in a flow of oxygen in order to avoid lattice
damage by thermal effects. The products were characterized in the fully rehydrat-
ed state. Deammoniation was monitored by IR spectroscopy but not quantitative-
ly evidenced by chemical analysis (e.g., Kjeldahl titration). The products were not
subjected to a subsequent heat treatment at low temperatures. Thus, this study is
irrelevant to the thermal stability of H-Y zeolite in the presence of intracrystalline
water. Nevertheless, the crystallinity of the rehydrated high-exchanged sample
was found to deteriorate even at ambient temperature, while the low-exchanged
H,Na-Y zeolite proved to be not affected.
In any case, rehydrated H-Y zeolite is a highly delicate material the structure
of which may be, depending on the exchange degree and the framework Si/Al
Dealumination Techniques for Zeolites 211

ratio, more or less affected at low or even ambient temperature by the intrinsic
acidity. Nevertheless, the influence of intracrystalline water on the structure of
hydrogen zeolites is a phenomenon not always considered in experimental prac-
tice and interpretation of findings. Results obtained by even highly sophistica-
ted techniques can be only unreservedly accepted, especially in case of sample
pretreatments involving a heating step, if the deammoniation was performed in
situ in the measuring cells or if the pretreated samples were transferred to the
measuring cells under complete exclusion of atmospheric moisture and, of
course, if the measurements themselves were performed in absence of water.
This problem is well illustrated by the classical debate between Skeels and Kerr
[4547] on the existence of the hydrogen form after thermal decomposition of
NH4-Y. The author of the present review is convinced of the soundness of Kerrs
reasoning in favor of the existence of H-Y and its formation by deammoniation
of ammonium Y zeolite. Naturally, strict conditions must be observed in order
to prevent H-Y coming into contact with water and from thermal dehydroxyla-
tion. Thus, Kerrs decisive argument concerning the destabilization of the frame-
work of H-Y contacted for ion exchange with an aqueous salt solution is surely
sound. Even the counter-argument in [48] based on the 27Al MAS NMR spec-
troscopic detection of some octahedrally coordinated aluminum in deammo-
niated Y zeolite is not conclusive since the paper referred to does not give any
information whether the separately prepared H-Y was transferred to the sample
holder (rotor) under complete exclusion of atmospheric moisture and whether
dehydroxylation had been completely avoided.

2.4
Gaseous Halogen Compounds as Dealuminating Agents

Dealumination of zeolites with Si/Al ratios >5 by reaction with gaseous chlorine
compounds at elevated temperatures was first reported in a patent applica-
tion filed in 1975 [49]. According to the presented procedure, highly dehydrat-
ed zeolites were contacted, at temperatures higher than 400C, with Cl2 and/or
HCl or with a mixture of Cl2 and CO, preferably in a molar ratio of 1:1 as in
phosgene. The mechanisms of the reactions between these agents and frame-
work aluminum were not treated in detail; it was only suggested that first de-
fects were created which may be filled up, in a consecutive process, by migrating
silicon atoms.
Fejes et al. [5054] extensively studied the extractive dealumination of zeo-
lites (mordenite) with acid halides (phosgene, nitrosyl chloride) at 400600C.
The reaction of H-mordenite with phosgene, monitored by IR spectroscopic
determination of the volatile reaction products HCl and CO2 , was found to pro-
ceed in three main steps at temperatures above 100C (Eq. 2a) and 300C
(Eqs. 2b and 2c):
{AlO4/2} M+ + COCl2 {AlO4/2} C+ OCl + MCl , (2a)
{AlO4/2} C+ OCl {} + AlOCl + CO2 , (2b)
AlOCl + COCl2 AlCl3 + CO2 (2c)
212 H.K. Beyer

where M denotes Na or H and {} a lattice vacancy left by removal of one Al


and two O atoms from the framework.Aluminum chloride is volatile at the reac-
tion temperature and is purged out from the sample if the process is carried out
in a stream of phosgene.
The dealumination degree can be controlled by the reaction temperature.
Also, the sodium form of zeolites can be used as starting material; however, com-
parable dealumination degrees can be achieved, if at all, only at significantly
higher temperatures. NaCl, formed in this case as reaction product, remains in
the sample and has to be washed out after the dealumination reaction. The crea-
tion of structural vacancies was found to be associated with the appearance of
new IR bands of lattice vibrations at 930 and 860 cm1 for mordenite and fauja-
site-type zeolites, respectively. The dealumination technique reported was found
to result in only minor crystallinity losses (less than 10% [53]) when applied to
zeolites with relatively high Si/Al ratios (mordenite, clinoptilolite). Faujasite-
type zeolites could be subjected to this procedure without significant loss in
crystallinity only after preceding ultrastabilization [52]. Further publications
from Fejess group dealt with the mechanism of the dealumination with phosge-
ne studied by IR spectrometric and thermoanalytic techniques [54, 55], structu-
ral consequences [55], adsorption behavior and catalytic properties of thus-
modified mordenite [56].
Gaseous CCl4 and CHCl3 were found to react similarly with the framework
aluminum of mordenite [55]. In the case of carbon tetrachloride, the reaction
proceeds via phosgene formed as an intermediate, in analogy to the reaction
step shown in Eq. (2b), according to Eq. (2d) [55] while, according to [57], CHCl3
gives CO and HCl in a reaction step analogous to Eq. (2c).
{AlO4/2}C+Cl3 {} + AlOCl + COCl2 (2d)
In the environment of the vacant sites, structural rearrangements were believed
to proceed. However, this framework reconstruction and the chemical nature
of the vacancies denoted in Eq. (2b) by the not very instructive sign {} are
not yet really understood. The removal of one aluminum and two oxygen atoms
from the framework according to Eqs. (2a)(2c) should result in at least local
lattice changes (or damages) similar to those occurring upon dehydroxylation of
H-zeolites (see Sect. 2.1). Recently, the formation of SiCl (and some =SiCl2)
groups in Na- and H-mordenite upon dealumination with phosgene has been
suggested since the respective silanol groups were detected in rehydrated pro-
ducts by 29Si MAS NMR spectroscopy [57, 58]. The amount of silanols was found
to be close to two per extracted aluminum, thus, the following reaction was
believed to proceed:
| |
Si Si
| |
O H+ Cl
| | | | |
SiOAlOSi + 3COCl2 SiCl Si + HCl + AlCl3 + 3CO2 (3)
| | | | 
O O
| 
Si Si
| |
Dealumination Techniques for Zeolites 213

The postulation of SiCl species is undoubtedly an original element in the inter-


pretation of the chemical nature of vacancies in zeolites dealuminated with
phosgene even if the structural consequences of the severe lattice constraints
necessarily associated with the simultaneously formed new Si-O-Si bridge are
not considered. However, the crucial point is that the detection of SiOH (and
=Si(OH)2) groups in the washed products is not straightforward evidence for
the existence of SiCl precursors. The silanol groups could also have been for-
med by hydrolysis of strongly strained Si-O-Si bonds in defect sites created
during the preceding dealumination process. Unfortunately, no attempts were
made [57, 58] to detect and quantify by simple titration the HCl that must have
been released when the solid product of Eq. (3) was brought into contact with
liquid or gaseous water. Even mixed chlorine-fluorine nests with four SiCl(F)
groups have been suggested by the same authors to be formed upon dealumina-
tion of zeolites with CCl2F2 [59, 60]. However, as long as the simple proof of
existence of incorporated halogen is not forthcoming, the suggestion of nests
(vacancies) comprising two (Eq. 3) or four SiCl(F) groups is rather insubstan-
tial.

3
Hydrothermal Dealumination of Zeolite Frameworks

3.1
Early Fundamental Investigations

In 1967, McDaniel and Maher [61] reported a method to increase the thermal
stability of Y zeolite. This so-called ultrastabilization procedure immediately
became a matter of considerable interest because of the technical importance of
Y zeolite as a catalyst. The process consists of two major steps, (1) the practical-
ly complete removal of sodium ions by a two-step ammonium ion exchange with
intermittent heating and (2) the conversion of this material by heat treatment at
800C or above to a faujasite-type zeolite resistant to the influence of heat up to
about 1000C. Although McDaniel and Maher noticed that ultrastabilization is
associated with a decrease of the ion-exchange capacity and unit-cell size, the
question of the framework aluminum content of modified Y zeolite was not
explicitely raised in their paper.
The credit for recognizing the fundamental features of the stabilization
mechanism must go to Kerr [62, 63]. He evidenced that the water formed by ther-
mal dehydroxylation of the hydrogen form plays a decisive role in this process
and went on to state that any technique for keeping this water in the system
during the heating process will result in a stable product [62]. In line with this
statement it was shown [63] that thermal treatment of NH4-Y in thin (only a few
mm thick;shallow bed) layers resulted in thermally less stable products. In con-
trast, treatment in deep bed, i.e., in bed geometries impeding the fast removal
of the reaction product water from the bed by diffusion, yielded ultrastablezeo-
lites. Kerr suggested a stabilization mechanism comprising (1) hydrolytic cleava-
ge of -O-Al-O- bonds by self-steaming, i.e., contact with gaseous water formed
214 H.K. Beyer

as reaction product, (2) release of hydroxyaluminum species from the framework


and (3) occupation of cation sites by cationic aluminum species [62, 63]. Similar-
ly, ultrastabilized products were obtained also by direct steaming [64, 65], i.e., by
contacting the hydrogen zeolite during the thermal treatment with water steam
from external sources at partial pressures generally up to 1 bar.
Further evidence for removal of aluminum and oxygen from the framework
of Y zeolite ultrastabilized according to [61] was found in an X-ray study by
Maher et al. [66]. Moreover, it was shown that ultrastabilization is also associat-
ed with the incorporation of silica, originating from other portions of the
crystal, into the framework vacancies left by dealumination, and that this pro-
cess is an absolutely necessary step of the stabilization process. Based on IR
results, the same interpretation was given by Scherzer and Bass [67]. Gallezot et
al. [68] arrived at similar conclusions in an X-ray study of Y zeolite dealuminat-
ed with H4EDTA according to Kerr [24] since, even after extraction of 53% of the
framework aluminum, all tetrahedral sites and oxygen positions in the frame-
work were found to be completely occupied.
Later, as MAS NMR spectrometers became commercially available, impres-
sive evidence for the release of framework aluminum and the refilling of frame-
work vacancies by framework silicon atoms in hydrothermally treated Y zeolite
was derived from 27Al and 29Si spectra [48, 69, 70]. Figure 3 shows the 29Si MAS

Si(2Al)
Si(1Al)

Si(3Al)
Si(0Al)

a
Si(1Al)

Si(2Al)
Si(0Al)

Si(3Al)

bb
Si(0Al)

Si(1Al)

Si(2Al)

cc
80 90 100 110 120 80 90 100 110 120

ppm from TMS

Fig. 3ac. Experimental (left column) and computer-simulated (right column) 29Si MAS NMR
spectra of NH4 ,Na-Y zeolite a prior to and b after thermal treatment at 400C and c after
steaming at 700C [48]
Dealumination Techniques for Zeolites 215

NMR spectra of NH4(0.75)Na(0.25)-Y (a) prior to and (b) after thermal treatment at
400C and (c) after steaming at 700C [48]. The experimental spectra are given
in the left-hand column and the respective computer-simulated spectra with
deconvoluted individual signals (dotted lines) based on gaussian peak profiles
in the right one.
Upon steaming, spectrum (c) was completely transformed with respect to
that of the parent sample. The strong intensity increase of the Si(0Al) signal in
spectrum (c) at the expense of the signals typical of framework Si connected via
O atoms with 1, 2 or 3 Al indicates the replacement of Al by Si in the framework.
The much less pronounced changes in the intensity distribution after heat treat-
ment at 400C (spectrum b) point to an only slight release of framework alumi-
num probably not induced by thermal dealumination, which is basically asso-
ciated with the dehydroxylation of hydrogen zeolites. Since the applied pretreat-
ment temperature of 400C is high enough for the deammoniation of NH4-Y, but
not sufficiently high for the dehydroxylation of the resulting hydrogen form, it
seems to be obvious that no precautions were taken to avoid the resorption of
atmospheric humidity after the heat treatment and, consequently, the frame-
work was affected by the processes described in Sect. 2.3.
Scheme 1 reflects the so-called T-jump mechanism (T stands for the tetra-
hedrally coordinated framework atom) proposed by von Ballmoos [71] for this
hydrothermal healing process.
For nearly a decade not much stress was put on the question whether silicon
atoms, known to refill upon self-steaming or steaming the lattice vacancies cre-
ated by release of framework aluminum, originate from amorphous silica impu-
rities, from the surface of the zeolite crystallites, or from crystalline areas inside
the crystals. In 1980, Lohse et al. [28] showed in a study dealing with the adsorp-
tion of nitrogen and some hydrocarbons that a secondary pore system (about
0.13 cm3/g) with pore diameters between 3 and 3.8 nm was formed when NH4-
Y zeolite was subjected to steaming and subsequently extracted with hydro-
chloric acid. They evidenced that silicon, which reoccupied empty tetrahedral
sites, did not (or not exclusively) come from the surface of the crystals, but pre-
dominantly from the bulk, probably involving the elimination of entire sodalite
units. Essentially based on results of adsorption measurements, further evidence
for the formation of mesopore systems upon steaming of hydrogen zeolites was
presented for, e.g., Y zeolite [7274] and mordenite [23, 75]. Using both N2
adsorption and electron microscopy [76], the mesopore structure in steamed Y
zeolite was found to be best described by cavities of spherical shape of about
15 nm in diameter connected by narrower openings. A significant proportion of
the crystallite surface was covered by an amorphous layer.
It is evident that the aluminum released from the framework upon hydro-
thermal treatment remains in the sample, either as an intracrystalline oxidic or
cationic aluminum species or as intercrystalline material, i.e., as a separate
crystalline or amorphous aluminum oxide phase [7779]. Typical signals
appear in the 27Al MAS NMR spectra of Y zeolites upon hydrothermal dealu-
mination (Fig. 4). A line at 0 ppm associated with octahedrally coordinated
aluminum is indicative of hydrated cationic species [7982]. Amorphous oxi-
dic aluminum species, which are subject to large second-order quadrupolar
216 H.K. Beyer

Scheme 1. T-jump mechanism after von Ballmoos [70]

interactions, are revealed by a broad signal over a wide chemical shift range.
A line at about 30 ppm typically found in 27Al spectra of hydrothermally
treated zeolites is generally believed to be associated with five-coordinated
extra-framework aluminum. More recently, the line at 30 ppm was attributed
to tetrahedrally coordinated aluminum species with a sufficiently large qua-
drupole coupling to cause significant second-order shifts [83]. In any case,
no attempts were made to interpret in more detail the chemical nature of such
species.
In order to prepare high-silica zeolites or even pure silica with zeolite struc-
tures, techniques combining hydrothermal treatment and acid leaching were
applied. Generally, aluminum was first removed from the framework by steam-
ing or self-steaming resulting in acid-resistant zeolites, and then the extra-frame-
work species were extracted in a second step by leaching with acids. High-
Dealumination Techniques for Zeolites 217

a
200
200 100
100 00
0 -100
100

b
200
200 100
100 00
0 -100
100
ppm from Al (H2O)63+

Fig. 4a, b. 27Al MAS NMR spectra of hydrothermally dealuminated Y zeolite after a partial and
b complete rehydration [79]
218 H.K. Beyer

ly aluminum-deficient mordenites [23, 84, 85], faujasites [8688], offretites


[8991] and ZSM-5 [10, 92, 93] were prepared in this way.
In some instances the materials were also repeatedly subjected to this two-
step treatment to enhance the degree of dealumination, e.g., in [89]. In the case
of zeolites resistant to acids due to high framework Si/Al ratios, e.g., mordenite
and ZSM-5, acid leaching may also precede the hydrothermal treatment [10, 23].
In this case, healing of framework hydroxyl nests left by aluminum release is the
lattice-stabilizing process. The hydrothermal stability of steamed Y zeolites was
found to increase upon removal of extra-framework aluminum by extraction
with acids [94].
Using X-ray photoelectron spectroscopy, Gross et al. [95] studied the surface
composition of Y zeolites dealuminated by hydrothermal treatment and by
extraction with EDTA according to [24]. In hydrothermally dealuminated sam-
ples, it was found that remaining aluminum accumulated at the outer crystal
surface of zeolites in the form of oxidic clusters, but not of a dense layer. The
existence of cationic aluminum species could not be confirmed. Extraction of
Y zeolite with EDTA favored the dealumination of the external crystal shell,
probably resulting in an aluminum concentration gradient along the radii of
the crystals [95].
It must be emphasized that zeolites can be hydrothermally dealuminated only
in their hydrogen form or in cation forms convertible into the hydrogen form
upon thermal treatment. Consequently, the dealumination degree is limited by
the efficiency of the ammonium or proton ion exchange generally preceding
the dealumination procedure. For example, steaming of Na-ZSM-5 at 700C
for 18 h resulted in the transformation into cristobalite, whereas under the
same conditions the hydrogen form of that zeolite retained its crystal topology
and the framework aluminum content decreased from 1.17 to 0.25 Al per unit
cell [93].

3.2
Review of Recent Investigations

3.2.1
Faujasite-Type Zeolites

A series of papers [9699] have dealt with the influence of the operating condi-
tions of the hydrothermal dealumination technique (steaming temperature and
time, water partial pressure, flow rate of the steam, etc.) on the controlled pre-
paration of zeolitic materials with improved and optimized physical-chemical
properties, especially for application as catalysts. Upon steaming at 500C, de-
alumination of Y zeolite proceeded in two steps, a fast one up to a dealumination
degree of about 50% in the first 30 min and a second slow one resulting in
further progressive dealumination [97]. From the results of 27Al MAS NMR spec-
troscopy it was concluded that during the first step octahedrally coordinated
extra-framework aluminum species were formed, while the aluminum expelled
during the second step was trapped in the sodalite units in a tetrahedral envi-
Dealumination Techniques for Zeolites 219

ronment. As to the two-step dealumination, Wang et al. [99] experienced a simi-


lar behavior of Y zeolite upon steaming at 820C. For both periods the kinetic
order with respect to water was found to be equal to 1. The first hydrothermal
step was suggested to correspond to the release of framework aluminum asso-
ciated with bridged hydroxyls (protons), while the second one (acid leaching)
was believed to proceed in zeolite regions where cationic aluminum species
compensate negative charges due to residual framework aluminum. Aluminum
was found to enrich at the external crystal surface only upon steaming at high
water pressure (>50 kPa) and never during self-steaming [99]. Physicochemical
characteristics of the dealuminated zeolite samples prepared in this way and
their catalytic behavior in the cracking of n-heptane were also reported [100].
Lutz et al. [101] also studied the chemical nature of extra-framework aluminum
formed upon steaming and concluded that this process results, concomitantly
with the creation of a mesopore system, in the formation of an X-ray amorphous
aluminum aluminosilicate on the external crystal surface and of highly conden-
sed Al cations compensating the residual negative charge of the framework.
Macedo et al. [102] arrived at similar conclusions in an IR and ammonia TPD
study of extra-framework aluminum species in steamed Y zeolites unleached
and leached with hydrochloric acid. The acidity in steamed and subsequently
acid-leached Y zeolites and mordenites was investigated by calorimetric deter-
mination of differential heats of ammonia adsorption [103]. Recently, samples
obtained by steaming of Y zeolite at 600C for different times were characteri-
zed with respect to acidity, aluminum distribution and adsorption behavior and
tested as catalysts for pentane conversions [104]. Dealuminated HY samples
were obtained by varying the steaming time between 16 and 98 h and the tem-
perature between 200 and 500C [105]. The products, tested as catalysts in 2-
methylpentane cracking, showed good X-ray crystallinity up to steaming tem-
peratures of 400C.
Li et al. [106] removed aluminum from NH4Y at 95C with an aqueous solu-
tion of oxalic acid and ammonium oxalate prior to heat treatment at 600C
under typical self-steaming conditions. The product with a framework Si/Al
ratio of 2.95 was then subjected to acid leaching with 1 M sulfuric acid followed
by a second self-steaming treatment. The final zeolite, used as catalyst for al-
kylation of phenol with long-chain olefins, retained 90% of the initial crystallin-
ity at a final Si/Al ratio of 5.71 and had a catalytically favorable mesopore volume
of about 0.3 ml/g.
Zhixiang et al. [107] compared typical characteristics of aluminum-deficient
Y zeolites prepared by steaming followed by acid leaching and produced by the
reverse sequence of the two steps. The sample obtained by the reverse method
showed, in contrast to the other one, aluminum enrichment on the external sur-
face of the crystals and lower charge and acid site density on the intracrystalli-
ne and mesopore surface.
Using X-ray photoelectron spectroscopy, Corma et al. [108] suggested a
migration of non-framework aluminum to the surface of the zeolite crystals
during deep-bed calcination of NH4Y zeolites. However, the reported high sur-
face Si/Al ratio, based on the Si 2p XPS signal and an Al 2p line at 73.8 eV assig-
ned to extra-framework aluminum species, seems to contradict this suggestion.
220 H.K. Beyer

On the other hand, the intensity of an Al 2p line at 75.0 eV attributed to tetrahe-


drally coordinated aluminum is indicative of a gradient in the framework alu-
minum content, this value being highest at the surface. In another paper [109]
these authors compared amount and distribution as well as the catalytic
cracking behavior of extra-framework aluminum species in Y zeolites dealumi-
nated by steaming combined with citric acid leaching and treatment with SiCl4
and (NH4)2[SiF6].
IR bands at 3693 and 3606 cm1 and the signal at 2.6 ppm in the 1H MAS NMR
spectrum were found to be typical of non-framework aluminum species in hy-
drothermally treated Y zeolites [110]. A re-investigation of the hydroxyl stretch-
ing range of IR spectra of hydrothermally treated and acid-leached Y zeolites
[111] essentially confirmed the assignments of bands in earlier papers. A com-
prehensive description and assignment of the 27Al and 29Si MAS NMR signals in
spectra of steamed Y zeolites were presented in [112].
Based on results obtained with the X-ray radial diffusion technique, Shannon
et al. [113] suggested that, in steamed Y zeolite, non-framework aluminum is
present in the form of boehmite-like clusters occluded in supercages or in the
mesopores formed during the hydrothermal healing process, but not located on
the external crystal surface or, as a separate phase, in the intercrystalline space.
To support this hypothesis, these authors relied on similarities in the OH-
stretching IR spectra of hydrothermally dealuminated Y zeolite and pseudo-
boehmite along with similar heats of ammonia adsorption.
The influence of acid and basic treatments on Y zeolites dealuminated to
various degrees was investigated by Aouali et al. [114]. The crystal structure of
low dealuminated samples was found to be unstable in acidic media, but did not
show drastic modifications in basic solutions. In contrast, Y zeolites highly de-
aluminated by combined hydrothermal treatment and acid leaching or by SiCl4
(see Sect. 4.1) proved to be highly resistant to acids but easily lost their crystal-
line structure in basic solutions. In another paper [115], the stability of ultra-
stabilized Y zeolites towards steaming at 750 and 810C was the subject of in-
vestigation.
He et al. [116] and Wan and Shu [117] reported on the influence of calcination
and hydrothermal treatment on compositional characteristics and thermal sta-
bility of rare earth containing Y zeolites and their performance in catalytic
cracking. The alkaline and hydrothermal stability of Y zeolites dealuminated via
hydrothermal treatment and by the SiCl4 technique was studied by Lutz et al.
[118]. Hydrothermal treatment was found to increase the chemical resistance of
Y zeolite to superheated water at 200C as well as to alkaline solutions due to the
formation of a protective layer of extra-lattice oxidic aluminum species on the
external surface of the zeolite crystals. The removal of this layer by acid leaching
resulted in significantly less stable products.
Sulikowski [29] calculated the fractal surface dimension of NH4,Na-Y zeolite
steamed at 550C from adsorption data of short-chain alcohols. This property of
matter being a quantitative measure of surface roughness was found to be 2.25,
which points to a significant deviation from a smooth surface due to the forma-
tion of extra-framework aluminum species and mesopores at the external sur-
face.
Dealumination Techniques for Zeolites 221

Fig. 5. Effect of steaming severity (partial pressure of water) on the hexane cracking activity
of H-ZSM-5 [119]

3.2.2
ZSM-5
In 1986, Lago et al. [119] reported on the progressive dealumination of the
framework of H-ZSM-5 upon hydrothermal treatment at 540C in a gas stream
with increasing partial pressures of water vapor up to 90 kPa. As illustrated in
Fig. 5, mild steaming was found to create sites with a catalytic activity for hex-
ane cracking significantly greater than that of the normal acid sites in ZSM-5.
The model suggested for the sites of enhanced activity (acidity) involves the
transformation of one framework aluminum of a structural unit, comprising
two paired tetrahedrally coordinated Al atoms, by partial hydrolysis to the hexa-
coordinated state. This species acts as an electron-withdrawing center and,
hence, increases the strength of the Brnsted acid site associated with the second
aluminum atom of the original pair that remains tetrahedrally coordinated.
A more detailed model (Scheme 2) of the modification of Brnsted acid sites
of the Al-(OH)-Si-O-Al-(OH)-Si type in an acidic zeolite upon mild steam-
ing was proposed in [120]. It comprises:
1. the conversion of a framework fragment of a Brnsted acid site from the ini-
tial state (I) into configuration (II) by partial hydrolysis,
2. the rearrangement of (II) to a first type (III) and then by subsequent dehy-
dration to a second type (IV) of strong Brnsted acid sites (positive partial
charge, d+, on the respective oxygen atom of a Al-(OH)-Si configuration)
and, finally,
3. the formation of cationic extra-framework aluminum species (e.g., AlO+)
upon more severe treatment (higher temperature and/or H2O partial pres-
222 H.K. Beyer

Scheme 2. Proposed model of the modification of Brnsted acid sites of the Al-(OH)-Si-O-
Al-(OH)-Si

sure and/or prolonged reaction time) resulting in a complete separation of


Al ions from the framework, after Khl [121].
H-ZSM-5 zeolites were found to dealuminate upon steaming at 500600C,
independently of the crystallite size, according to a second-order kinetics with
respect to the aluminum concentration of the starting zeolites [122]. Though it
was not explicitly stated, data given in the paper allows the conclusion that the
obtainable minimum aluminum level amounts, independently of the starting
Si/Al ratio, to about 0.3 Al per unit cell.
Campbell et al. [123] investigated the effect of hydrothermal treatment at a
water vapor pressure of about 2 kPa and at 600840C on the structure and
adsorption properties of, and on the nature and distribution of aluminum spe-
cies in, ZSM-5 zeolites with 0.777.6 Al per unit cell in the framework of the
starting materials.
Dealumination Techniques for Zeolites 223

During steaming of H-ZSM-5 at 550C two processes were found to influence


the concentration of Lewis acid sites in opposite directions, (1) the dealumina-
tion resulting in extra-framework aluminum species and (2) the agglomeration
of this species associated with a decrease in Lewis acidity [124]. The significan-
ce of this observation for the use of steamed zeolites as catalysts was discussed.
Depending on the steaming conditions, up to five different types of acidic sites
assigned either directly to extra-framework aluminum species or to their inter-
action with silanols were found in hydrothermally treated H-ZSM-5 by diffuse
reflectance infrared spectroscopy [125].
In zeolites of the ZSM-5 type prepared by recrystallization of rare earth con-
taining Y zeolite, hydrothermal dealumination was found to be suppressed, and
crystallinity to be better retained by the rare earth constituent [126].

3.2.3
Other Zeolites
Hydrothermally dealuminated H-EMT zeolites were prepared by steaming in
the temperature range of 450740C; resulting materials both unleached and
leached with 1 N HCl were characterized [127].
In a series of mordenites dealuminated, prior to calcination at 550650C, to
various degrees by acid extraction, Musa et al. [128] experienced non-mono-
tonic variations of several characteristics (unit cell and crystallite size, Lewis and
Brnsted acidity, bulk/surface Al concentration ratio) which were explained by
a different dealumination behavior of crystallites differing in size and crystal
symmetry. Small crystals having preferentially Cmmm symmetry were assumed
to undergo even under mild conditions a fast dealumination distorting the
structure and resulting in fragmentation of the crystallites or, under more severe
conditions, in partial amorphization. In contrast, crystal structure and size of
larger mordenite particles with preferential Cmcm symmetry are believed to be
highly resistant to even severe dealumination.
Physicochemical and catalytic properties of mordenites dealuminated by
both separate and combined thermal treatment and acid leaching and by com-
bined thermal/hydrothermal treatment have been characterized by various
techniques [129]. Dry heat treatment at 750C expelled as much as 70% of the
aluminum from the framework resulting for ammonia in nearly complete inac-
cessibility of strongly acidic sites [130]. Indication of acidity was partly restored
upon acid leaching, which showed that sites associated with residual framework
aluminum again became accessible to basic molecules by extraction of the
extra-framework aluminum species. Roughly 40% of the aluminum remaining
after heat treatment at 750C in the framework was found to be expelled upon
steaming at 500C. In line with the behavior of Y zeolite [99], steam was found
to be necessary for extra-framework aluminum species to migrate to, and ac-
cumulate at, the crystal surface [131].
Lee and Ha [132] treated Na-mordenite (SiO2/Al2O3 weight ratio = 6.5) with a
6 M HCl solution at 90C for 316 h or a 0.5 M HF solution at room temperature
for 192528 h. HCl was applied in large excess, HF in amounts about equivalent
to the aluminum in the sample subjected to acid leaching. After washing with
224 H.K. Beyer

water, the acid-leached samples were steamed at 600 and 500C, respectively. The
HCl/steaming procedure resulted in products, the framework Si/Al ratios of
which increased with increasing leaching time. In contrast, in the case of HF
leaching, a preferential release of aluminum, causing an increase in the frame-
work Si/Al ratio up to about 20, was found to proceed only in the initial period,
while later simultaneous removal of Si and Al occurred. Disregarding samples
subjected in the first step for longer times to HF leaching, all products prepared
by HCl and HF leaching are claimed to retain more than 80% crystallinity and
to have also mesopores with diameters around 3.7 nm.
Fernandez et al. [133] adapted the combined technique consisting of hydro-
thermal treatment and successive extraction with acids, already successfully
used for the dealumination of Y zeolite [86, 87], for the dealumination of
K,TMA-offretite. With an increasing number of treatment cycles (up to 4), the
aluminum content could be progressively decreased from 3.41 to 1.68 Al per u.c.
The process was found to be associated with a significant increase in the ther-
mal stability of the lattice, but was also accompanied by the formation of defects
and holes in the crystal (mesopore system).
Carvalho et al. [134] reported on the dealumination of a synthetic K,TMA-
offretite by two succesive thermal treatments under self-steaming conditions at
550 C and higher temperatures up to 850 C with intermittent removal of potas-
sium cations by ammonium ion exchange. The dealumination, approaching a
degree of about 60%, was associated with a strong decrease in crystallinity.
Aluminum located in crystallographically different framework sites was found
to be removed to different degrees with preference for T2 sites. (As to the loop
configurations of T1 and T2 atoms in the structure of offretite, see [135].) As in
the case of the solid-state dealumination of the structurally related zeolite L with
(NH4)2[SiF6] (see Sect. 4.4 and [198]), only the cell parameter a was significant-
ly reduced by removal of framework aluminum. Offretite behaved similarly to Y
[99] and mordenite [129] inasmuch as no migration of extra-framework alumi-
num species to the outer surface of the zeolite crystals was observed under the
applied self-steaming conditions. Contrary to statements in [134], a dealumina-
tion preference was found for T1 sites of the offretite lattice in a study dealing
with MAS NMR spectroscopic characterization of steamed offretite and erionite
[136].
The behavior of zeolite W (synthetic mazzite) was intensely studied by French
research groups [137141]. McQueen et al. [137] found large differences in the
size and volume of mesopores depending on the aluminum content of the parent
zeolite. Mesopores of 10 nm in diameter with a total mesopore volume of
0.050.06 ml/g were found in the dealuminated variety of the aluminum-rich
parent material, while dealumination of the aluminum-poor parent zeolite
yielded, at about the same final aluminum content corresponding to a Si/Al ratio
of about 22, mesopores with diameters of 68 nm representing a volume of
0.030.04 ml/g.
Massiani et al. [138] studied the effects of calcination at 500 C, subsequent
hydrothermal treatment (self-steaming) at 600C and, finally, acid leaching on
the structural parameters, porosity and acidity of zeolite W. Removal of 25 and
50% of the framework aluminum, preferentially from sites located in the four-
Dealumination Techniques for Zeolites 225

membered rings of the gmelinite cages, was observed upon calcination and self-
steaming, respectively. The consequences of the dealumination were basically
similar to those experienced with Y zeolite. Chauvin et al. [141] subjected zeolite
W with varying residual sodium contents to hydrothermal treatment at 500C
and subsequent acid leaching at 80C. Steaming resulted in the formation of
mesopores of 10 nm in diameter that were not available to sorbents due to depo-
sits of debris with aluminum in tetrahedral and octahedral configuration. Acid
leaching removed all the tetrahedrally and part of the octahedrally coordinated
aluminum species. The mesopores were found to be interconnected by narrow
necks and not directly linked to the crystallite surface. The influence of sequen-
tial steaming and acid leaching on the texture of synthetic mazzite (zeolite W)
has also been studied [141].

3.3
Thermal Dealumination

In the description of the preparation of ultrastabilized Y zeolite reported in [61]


it was not explicitly stated that water vapor is involved in the dealumination pro-
cess. Also, in numerous other papers dealing with thermal dealumination
and/or thermal stabilization of zeolites, the role of water was ignored and the
stabilization process was assumed to be governed by a purely thermal effect.
As is well known, thermal treatment of ammonium zeolites to about 400
500C results in the release of ammonia (deammoniation) and in the formation
of the hydrogen form of the respective zeolite. The hydrogen form of a zeolite,
especially of zeolite Y, is generally a highly delicate material inasmuch as it
undergoes dehydroxylation in a temperature interval slightly exceeding, but
sometimes also overlapping with, that of the deammoniation [142]. Moreover,
dehydroxylation is accompanied by a release of aluminum from the framework
[143] and obviously by a gradual collapse of the crystal lattice (amorphization),
as revealed by data reported, e.g., in [123, 130, 144]. In any case, the hydrogen
form of Y zeolite obtained by thermal decomposition of the ammonium form
was found to be significantly less thermostable than the sodium or other cation-
ic forms [51, 143]. In a series of systematic studies it was clearly shown that heat
treatment of hydrogen Y zeolites, cf. e.g., [63, 65], and mordenites, cf. e.g., [23,
145], alone did not stabilize zeolite lattices if the water, formed during the pro-
cess by dehydroxylation, was efficaciously removed. Thus, in studies claiming
dealumination and stabilization of zeolite frameworks by mere thermal treat-
ment (cf. e.g., [146]), obviously water evolved during the process as an intrinsic
component of the reaction system was involved in the observed dealumination
and/or stabilization by inducing the lattice healing process.
226 H.K. Beyer

4
Isomorphous Substitution of Framework Silicon for Aluminum

4.1
Dealumination with Silicon Tetrachloride

4.1.1
Faujasite-Type Zeolites
In 1980, Beyer and Belenykaja [147] reported for the first time on a process in
which zeolitic framework aluminum is directly replaced by silicon. The funda-
mentally new idea was to adapt the reaction between alumina and silicon
tetrachloride, Eq. (4), first reported by M. Daubre [148] nearly 150 years ago in
1854, also for the dealumination of zeolites, i.e., to use gaseous silicon tetra-
chloride as dealumination agent and, at the same time, as extraneous silicon
source.
2Al2O3 + 3SiCl4 4AlCl3 + 3SiO2 . (4)
It was found [147, 149] that, under appropriate experimental conditions, frame-
work silicon is directly isomorphically substituted for aluminum in a strongly
exothermic reaction without any substantial lattice damage when the sodium
form of faujasite-type zeolites was contacted with silicon tetrachloride vapor at
temperatures around 500C. In a subsequent study [150], the isomorphous
incorporation of Si was also evidenced by striking differences in 29Si MAS NMR
spectra. Before dealumination, the spectrum of Na-Y (Fig. 6, spectrum a)
exhibited the typical signals assigned to silicon atoms with 0, 1, 2 and 3 Al atoms
in the second coordination sphere. In contrast, the spectrum of Y zeolite subjec-
ted to nearly complete dealumination with silicon tetrachloride (Fig. 6, spec-
trum b) consisted only of the signal at 107 ppm indicative of Si(4Si, 0Al) orde-
ring. The high crystallinity and structural homogeneity of this material is indi-
cated by the extreme sharpness of this line and also reflected by the high-reso-
lution electron micrograph presented in Fig. 7, which is exactly the same as for
the parent Na-Y zeolite.
Only part of the AlCl3 , volatile at the reaction temperature, was found to esca-
pe from the zeolite bed during the dealumination of Na-Y. This behavior was
attributed to the formation of sodium tetrachloroaluminate [147]. Further, a
strong exothermic reaction not or scarcely resulting in framework dealumina-
tion was already observed at temperatures (about 250C) substantially lower
than those needed for isomorphous substitution and found to result in the
formation of surface SiCl3 groups revealed by a 29Si MAS NMR signal at
45 ppm [151]. Thus, the over-all process
{AlO4/2} Na+ + SiCl4 {SiO4/2} + Na[AlCl4] (5)
obviously comprises three major steps:
1. the reaction of SiCl4 with sodium cations resulting in surface SiCl3 groups
and NaCl
{AlO4/2} Na+ + SiCl4 {AlO4/2} SiCl3+ + NaCl, (5a)
Dealumination Techniques for Zeolites 227

Si(2Al)

Si(1Al)

Si(0Al)
Si(3Al)

aa
-40
40 -80
80 -120
120
d / p.p.m.
d / p.p.m.

bb
-40
40 -80
80 -120
120
d / p.p.m.
d / p.p.m.
Fig. 6a, b. 29Si MAS NMR spectra of Na-Y a prior to and b after treatment with SiCl4 at 560C
[150]
228 H.K. Beyer

a c
Fig. 7. a High-resolution electron micrograph of zeolite Y dealuminated with SiCl4 along the
[110] zone axis; b selected area diffraction pattern corresponding to this zone axis; and c [100]
projection of the framework [150]

2. the isomorphous replacement of framework aluminum


{AlO4/2} SiCl3+ {SiO4/2} + AlCl3, (5b)
3. the formation of sodium tetrachloroaluminate first described by Whler in
1827
NaCl + AlCl3 Na[AlCl4] (5c)
where {AlO4/2} and {SiO4/2} refer to primary tetrahedral units of zeolite struc-
tures containing Si and Al, respectively.
The chemistry of this dealumination technique may appear to be simple;
however, it is rather difficult and requires exact knowledge of this complex pro-
cess and great experience in its application to prepare completely and, especial-
ly, partially dealuminated Y zeolites without essential damage to the lattice.
Depending on the amount of applied zeolite, bed geometry and reaction con-
ditions, the strongly exothermic nature of reaction (5) can mean that consider-
able overheating may occur in the zeolite bed and result in a partial or total
collapse of the crystal structure [147, 149]. At reaction temperatures usually
recommended for that dealumination process (about 500C), Eq. (5c) is not yet
significantly shifted to the left side, i.e., most of the aluminum removed from the
framework is present in the form of sodium tetrachloroaluminate deposited in
the large cavities of the faujasite structure and only a minor part, depending on
reaction time and temperature, can escape in the form of AlCl3 via sublimation
Dealumination Techniques for Zeolites 229

[147]. A signal at 100.7 ppm found in the 27Al MAS NMR spectrum of NaY zeo-
lite treated with SiCl4 was reported to be indicative of Na[AlCl4] [152].
Unfortunately, it has not proved to be practicable to enhance the decomposi-
tion of the complex salt and, hence the escape of AlCl3 , simply by temperature
increase since the crystal structure of faujasite was found to be strongly affected
when contacted with SiCl4 at about 550C and even to be destroyed under some-
what more severe conditions [149]. Thus, in this dealumination procedure, the
increase in the reaction temperature is strictly limited. The amorphization of the
product at high reaction temperatures (550C and above) was also observed by
Anderson and Klinowski [152] who reasoned that the rate of aluminum removal
could be greater than that of silicon substitution at these temperatures. However,
this effect also appeared when the substitution was first carried out, as usual, at
about 500C and then the product was heat-treated at higher temperatures
[149]. In contrast, a highly thermostable product was obtained when the com-
plex salt was removed by washing. Thus, it was evident that occluded Na[AlCl4]
affects the crystal structure at temperatures of 550C and above and the extrac-
tion of the complex salt by washing with water seems to be the only promising
way to remove this reaction product from the dealuminated zeolite.
In practice [147], a stream of an inert gas saturated at ambient temperature
with silicon tetrachloride was passed at higher temperatures through a bed of
the zeolite, preferentially pelleted without any binder. To avoid overheating due
to released reaction heat, it was recommended to contact the zeolite with SiCl4
first at a lower temperature (about 200C), then to heat it in the gas stream con-
taining SiCl4 at a moderate rate up to the reaction temperature of about 500C
and to continue the treatment at this temperature, e.g., for 2 h. Prior to this
treatment, the zeolite must be completely dehydrated, e.g., in a stream of dried
inert gas in situ in the reactor at about 400C, to avoid hydrolysis of the reagent.
After washing with water, a near-silica analogue of faujasite with Si/Al ratios of
5060 was obtained.
Y zeolite dealuminated with SiCl4 proved to be, after removal of Na[AlCl4] by
washing, extremely heat resistant; structure collapse was found to start only at
about 1200C [147, 149].
Furthermore, it proved to be a typically hydrophobic adsorbent. The adsorp-
tion isotherms of organic compounds showed the rectangular shape with satu-
ration at very low pressures (see Fig. 8) typical of the adsorption on high-alumi-
na zeolites. In contrast, ammonia and especially water are much less or prac-
tically not adsorbed (isotherms (d) and (e) in Fig. 8). [See the striking difference
in the adsorption of water on Na-Y (dotted isotherm e) and dealuminated Y
(continuously traced isotherm e; Fig. 8).] This distinctly hydrophobic adsorp-
tion behavior of highly dealuminated faujasite indicates that the framework
must be virtually free of lattice vacancies necessarily associated with hydro-
philic internal silanol groups.
The isomorphous substitution of framework aluminum in faujasite-type zeo-
lites was evidenced by chemical analysis, XRD (unit cell contraction), mid-infra-
red spectroscopy (shift of lattice vibrations) and adsorption measurements
(hysteresis-free adsorption isotherms) [147] and later by 29Si MAS NMR spec-
troscopy [149, 150, 153] (see Fig. 5).
230 H.K. Beyer

e
0.3
0.3

a b

c
3 3 . 11

0.2
0.3
gg

Adsorbed liquid / cm3 g1


liquid
Adsorbed
Adsorbed / cm
/ cm
liquid

0.10
0.10

0.1
0.3

d 0.10
0.05

0.1 0.2 0.3 0.4


Relative pressure (p/p0)

Fig. 8. Adsorption isotherms on Na-Y (dotted lines) and Y zeolite (Si/Al = 44) dealuminated
with SiCl4 (continuous lines). a n-Hexane; b n-butane; c benzene; d ammonia; and e water
[147]

The degree of dealumination depended first of all not only on the final reac-
tion temperature but also on the reaction time [149]. To convey some ideas of the
influence of these parameters, maximum dealumination (Si/Al about 50) was
obtained at 450C in 40 min while after 15 min the dealumination degree cor-
responded, even at a higher reaction temperature (475C), only to a Si/Al ratio
of 19.Although the dealumination degree could be easily controlled by these two
parameters, other difficulties associated with the removal of the complex salt
arose in case of partial dealumination. The main problem was that the tetr-
Dealumination Techniques for Zeolites 231

achloroaluminate complex hydrolyzed during the washing process and the thus-
generated acidity might have leached out, depending on the actual aluminum
content of the framework, more or less aluminum that had remained in tetra-
hedral framework positions after the SiCl4 treatment. This undesired consecutive
dealumination reaction, which resulted in lattice defects (hydroxyl nests), con-
sumed acidity created by hydrolysis of Na[AlCl4]. Thus, at not too high degrees
of dealumination, the pH of the washing water may be higher than 3 and either
hydroxyaluminum species, e.g., Al(OH)2+, which occupied lattice cation sites, or
even intra-crystalline or extra-crystalline Al(OH)3 might have formed. The
extra-framework aluminum content was found to pass through a maximum at
intermediate extents of dealumination [149]. The chemical nature of such extra-
framework aluminum species created in Y zeolite by partial dealumination (up
to Si/Al = 20) with gaseous SiCl4 and subsequent washing with water has been
intensely studied [154].At least part of this extra-framework aluminum was pre-
sent as cationic species contributing to the compensation of the skeletal charge
and acting as strong electron-acceptor (i.e., Lewis acid) sites. The IR spectra of
dealuminated Y zeolites were found to differ from that of the parent zeolite. An
additional band at 3620 cm1 was assigned to amorphous extra-lattice alumi-
num oxide species, and the intensity increase of the band at 3740 cm1, typical of
non-acidic SiOH hydroxyls, was attributed to lattice defects remaining after dea-
lumination. The dealuminated Y samples contained bridged hydroxyls, the stret-
ching frequencies of which (3630 and 3560 cm1) were similar to, but never-
theless significantly different from, those present in the HY zeolite (3645 and
3550 cm1, respectively). Disregarding some slight deviations in hydroxyl stretch-
ing band positions, similar observations and assignments were made in [151]. It
was stressed that the band found at 3730 cm1 is associated with the hydroxyl
nests formed during the washing process by acid leaching since this band com-
pletely disappeared upon steaming. [There is ample proof (see Sect. 3) that steam-
ing anneals lattice vacancies of the hydroxyl nest type.] In contrast, the band
at 3750 cm1 developed in intensity as the crystallinity deteriorated. Garraln et
al. [155] were inclined to assign the band at 3610 cm1 in spectra of samples de-
aluminated with SiCl4 at 350400C to amorphous aluminosilicate formed from
oxidic extra-framework aluminium species and silica originating from SiCl4 .
Using both the Bloch Decay and crossed-polarization techniques in a 29Si MAS
NMR spectroscopic investigation of a Y zeolite practically completely dealumi-
nated with SiCl4, Ray et al. [156] detected, besides tetrahedrally coordinated
framework silicon as the main constituent, three different types of SiOH groups
assigned to defect sites and a small amount of amorphous silica. Steaming of the
sample at temperatures up to 700C resulted in the elimination of the defect sites
and the disappearance of the amorphous phase due to the well-known hydro-
thermal healing process.
The influence of both reaction temperature and washing conditions on the
amount and distribution of extra-framework aluminum species in Y zeolites
dealuminated with SiCl4 have been studied [157]. Increase in the reaction tem-
perature and efficiency of washing (water > ethanol > ethanol + buffer) resulted
in an increase in the Si/Al ratio both in the bulk and at the surface. Temperature
increase and milder washing enhanced the Al-enrichment of the surface with
232 H.K. Beyer

respect to the bulk. The higher aluminum content at the surface was due to non-
framework aluminum deposits.
An attempt was made to separately estimate the degree of dealumination
obtained in the first process step upon treatment with SiCl4 and during the sub-
sequent washing procedure for highly dealuminated Y zeolite (Si/Al = 50) [149].
Starting from a Y zeolite containing 54.5 Al per unit cell, 3.5 Al per unit cell were
found in the final product after dealumination and, based on thermogravimetric
data, it was concluded that four more Al atoms were leached from the unit cell
by the acidity created during the washing step. Thus, it may be concluded that
the isomorphous substitution stopped, due to product inhibition, at a level of
47 Na[AlCl4] per unit cell corresponding to about 6 per large cavity, provided the
amount of reaction product that escaped in the form of AlCl3 could be neglected.
For X zeolite (85 Al per unit cell) it was shown by quantitative volumetric
measurements [149] that at 210C the maximum consumption of SiCl4 amount-
ed to about 38 per unit cell.An increase in the reaction temperature to 530C did
not affect the SiCl4 uptake. Thus, the reaction stopped, probably due to product
inhibition, at a level of 38 Na[AlCl4] per unit cell (4.7 per large cavities) cor-
responding to a framework Si/Al ratio of 3.03. The final washed products proved
to be X-ray amorphous, obviously due to the relatively high aluminum content
which favored framework damage by strong mineral acidity created during the
washing process by hydrolysis of the occluded Na[AlCl4]. In contrast, Sulikows-
ki and Klinowski claimed [158] that the lattice destruction observed upon dea-
lumination of Li,Na-X was due to removal of aluminum from six-membered
rings containing three aluminum T atoms which are present in larger numbers
in the structure of zeolite X than in zeolite Y. However, since the presence of
occluded lithium and/or sodium tetrachloroaluminate was not ruled out by the
authors, it could be that the lattice collapse of the X zeolite occurred during the
washing process by acid leaching rather than during the SiCl4 treatment.
In any case, the degree to which both isomorphous substitution and acid
leaching contribute to the overall dealumination of Y zeolites obviously depends
on experimental details in a way difficult to control. In contrast to hydrothermal-
ly dealuminated Y zeolite, extra-framework aluminum could not be detected by
XPS at the surface of Y zeolite dealuminated with SiCl4. However, the framework
Si/Al ratio was found to be considerably lower (2.4) than that of the bulk (8.0)
[108]. Anderson and Klinowski [152] concluded from Lewis acid site levels,
found to be low in Y zeolites dealuminated to different degrees with SiCl4 , that
little extra-framework aluminum was present in the samples. Shi et al. [159]
measured the heat of adsorption of ammonia on a series of Y zeolites with about
35 to 3 framework Al per unit cell prepared by isomorphous substitution with
SiCl4 in the reaction temperature range 200450C. Extra-framework alumi-
num was found only in products which were refluxed in hydrochloric acid at
pH 2 immediately after the treatment with SiCl4, but not in those washed with
water. In another report [160] these authors even questioned the formation of
Na[AlCl4] and suggested that all aluminum removed from the framework es-
capes in the form of AlCl3 since, according to these authors, no non-frame-
work aluminum remained within the porous structure or at the surface even at
partial dealumination.
Dealumination Techniques for Zeolites 233

However, this statement was obviously derived only from 27Al MAS NMR spec-
troscopic results, and the bulk aluminum content was not considered. Since it is
well known [161] that part of the extra-framework aluminum may be invisible
for 27Al MAS NMR spectroscopy, it may be possible that intracrystalline or inter-
crystalline oxidic aluminum was overlooked. A recent paper by Stockenhuber
and Lercher [162] focused on the characterization of extra-framework species
and acid sites in Y zeolites subsequent to dealumination with SiCl4 . Extra-frame-
work aluminum species were found to be present after the treatment with SiCl4,
but partially extractable from the product with ammonia solutions. Besides this
type of species, a second one, rich in silica and located mainly outside the zeolite
channels, was also suggested. The effect of acid leaching subsequent to partial
dealumination of NaY zeolite with SiCl4 at 450C has also been studied [163].
Summing up, depending on the reaction temperature, the dealumination of
NaY zeolite with SiCl4 always results in an increase in the framework Si/Al ratio
due to direct isomorphous substitution. However, during the following washing
step associated with the creation of strong acidity, further framework aluminum
is leached out under formation of lattice vacancies and cationic and oxidic
extra-framework aluminum species. Under optimal reaction conditions, how-
ever, more than 90% of the original framework aluminum is removed by iso-
morphous substitution. Thus, dealumination with SiCl4 is the favored method if
highly siliceous faujasites have to be prepared and it is complementary to the
dealumination with (NH4)2[SiF6] in aqueous medium ([180182], reviewed in
Sect. 4.3), that may be more advantageous for the preparation of low-dealumi-
nated zeolites.
As shown, dealumination of NaY zeolite by SiCl4 is restricted by product inhi-
bition due to the formation and deposition of Na[AlCl4] in the zeolite cavities.
At first glance it may appear to be advantageous to start from the ammonium
form, since the respective ammonium complex salt, if it exists at all, completely
dissociates at the reaction temperature into its volatile components NH4Cl and
AlCl3. Such a modification was already considered in the first paper dealing with
this technique [147] and applied, but not studied in detail, by Hey et al. [163].
Recently, the crystallinity of NH4,Na-Y zeolite (60% NH4+ exchanged) de-
aluminated with SiCl4 at 545C was reported to be only 19% [164]. In our ex-
perience (unpubl. results), isomorphous substitution is accompanied by un-
desired, not yet fully understood, concomitant reactions resulting in partial loss
of the lattice integrity if the starting material is an ammonium Y zeolite. How-
ever, it has been shown that product inhibition can be avoided if the lithium-
exchanged form is subjected to the dealumination procedure [165]. Compared
with Na[AlCl4], the corresponding lithium complex dissociates at considerably
lower temperatures and the formed AlCl3 escapes almost completely during the
reaction at 500C. As a consequence, products with extremely high crystallinity
and structural homogeneity can be obtained, as illustrated by the 29SiMAS NMR
spectra of high-silica varieties obtained from Na-Y and Li-Y shown in Fig. 9.
Faujasites of extremely low aluminum content (Si/Al >200) and a minimum
of framework vacancies and mesopores can be obtained by acid leaching and
steaming of Y zeolite previously dealuminated with SiCl4 [147, 149, 166]. Even
large-scale production of dealuminated Y zeolite based on this combined proce-
234 H.K. Beyer

-107.5

-100.9 -104.6 -111.1

aa

bb

-95 -105 -115


Chemical shift, dTMS (ppm)
Fig. 9a, b. 29Si MAS NMR spectra of dealuminated faujasites prepared by treatment with SiCl4
from a Na-Y and b Li0.62Na-Y [165]
Dealumination Techniques for Zeolites 235

dure has been reported [167]. In a comparative study, Lutz et al. [118] reported
on the hydrothermal and alkaline stability of a commercially available high-sili-
ca variety of faujasite (Si/Al =150) and hydrothermally dealuminated Y zeolite
prepared in this fashion. High-silica Y zeolite (Si/Al about 150) prepared by iso-
morphous substitution with SiCl4 proved to be poorly resistant to steaming
[168]. The loss in lattice integrity was found to start already at temperatures of
about 120 C and approached the 100% value at about 180C. Hydrothermal
resistance was increased by covering the external surface of the crystals with a
layer of alumina or alkali aluminosilicates.

4.1.2
Other Zeolites
It is evident that the applicability of this dealumination process is limited to zeo-
lites with pore openings large enough to allow the penetration of the SiCl4 mole-
cule, and that the dealumination may be strongly or completely inhibited by
intracrystalline deposition of Na[AlCl4], especially in the case of zeolites with
one-dimensional or, with respect to the accessibility for SiCl4 molecules, quasi-
one-dimensional pore systems. Thus, it is not surprising that the dealumination
of ZSM-20, structurally closely related to faujasite, has been readily carried out
[169]. The framework Si/Al ratio of the lithium form of this zeolite, as obtained
from the ammonium form by solid-state ion exchange, could be increased by
contact with SiCl4 from 3.6 to >100; the aluminum content of the bulk was re-
duced to 0.24 mmol Al2O3/g corresponding to a Si/Al ratio of 34.
In the first paper dealing with this method [147] it was reported that attempts
to use this technique for the dealumination of L zeolite in its as-synthesized
K,Na-form and Na-mordenite failed. Later it was claimed [170] that synthetic
large-pore Na-mordenite is partially (24%) dealuminated with SiCl4 vapor at
700C. That is for this type of reaction an extremely high temperature and re-
sulted, at least in the case of faujasites, in complete destruction of the lattice.
Though it was stated that the treated material retained high crystallinity, the
crucial point was not pointed out in more detail, i.e., it was not evidenced that
the relatively low dealumination was not accompanied by a similarly slight loss
in crystallinity.
Namba et al. [171, 172] subjected H-ZSM-5 to silicon tetrachloride vapor at
temperatures between 450 and 650C and observed only a slight increase in the
bulk Si/Al ratio from 19 to 24, while the surface Si/Al ratio determined by XPS
increased, depending on the reaction temperature, from 18 to 39. Thus, the
external crystal shell was preferentially dealuminated upon contact with SiCl4 ,
obviously due to diffusion restrictions. In this way the contribution of the sur-
face layer to the catalytic activity of ZSM-5 zeolite could be diminished and the
shape-selectivity effect enhanced. In contrast, Thomas et al. [173] reported that
aluminum could be removed from the lattice of ZSM-5 by treatment with SiCl4
at 540C.
Na-L zeolite is much easier (i.e., at lower temperatures) to dealuminate than
its as-synthesized K,Na-form [174]. This has been attributed to the higher diffu-
sivity of SiCl4 in zeolites containing smaller lattice cations and, hence, that have
236 H.K. Beyer

a more open pore structure. Following essentially the procedure given in [147],
zeolite beta was dealuminated with SiCl4 at 450C without any noticeable
decrease in crystallinity to a level correponding to Si/Al = 39 [175].
Post-synthetic dealumination of the ammonium form of zeolite MCM-22 by
treatment with SiCl4 vapor at 450C according to [147] led to an increase in the
bulk Si/Al ratio from 11 to 20 [176]. The products were found to contain octa-
hedral aluminum detected by 27Al MAS NMR spectroscopy. It was found that
five crystallographically non-equivalent T-sites revealed in the structure of MCM-
22 by 29Si MAS NMR spectroscopy were affected to different degrees upon
dealumination with SiCl4 vapor.
The framework Si/Al ratio of zeolite W (synthetic mazzite) was increased by
treatment with SiCl4 at 500C from 4.24 to 6.00 without significant loss in
crystallinity [177]. The dealumination reaction was accompanied by a slight
increase in the hexagonal lattice parameter a while c remained unaffected. This
unusual phenomenon, i.e., cell expansion upon isomorphous substitution of
silicon for aluminum, as well as 27AL MAS NMR spectroscopic results, pointed to
a redistribution of aluminum on at least two crystallographically different
framework T-sites. 29Si MAS NMR spectra of offretite, erionite and zeolite W all
dealuminated with SiCl4 were presented in [178].
A partial dealumination of ferrierite increasing the bulk Si/Al ratio (deter-
mined by EDAX) from 4.6 to 7.0 was found upon treatment with SiCl4 at 550C
[152]. However, the few presented data did not evidence isomorphous substitu-
tion and do not give any structural information about the product. Considering
the small pore openings of the two-dimensional channel system of ferrierite,
it is probable that the observed dealumination is accompanied by a gradual lat-
tice collapse.

4.2
Isomorphous Substitution with Other Silicon Halides

Dealumination of Y zeolite was also performed with silicon chloroform, SiHCl3,


at reaction conditions usually used with the SiCl4 technique [147].The reaction
is obviously similar to that described by Eq. (5); however, it is not yet known in
detail. Special precautions must be taken to avoid the presence of any oxygen in
the reactant stream since SiHCl3 is flammable at the reaction temperature. The
only advantage offered by this reactant in comparison to silicon tetrachloride
may be its potential applicability to the dealumination of zeolites with smaller
pore diameters due to its smaller molecular size.
In a patent filed to the Union Carbide Corp. [179], the use of silicon tetraflu-
oride was claimed as dealuminating agent and extraneous silicon source. Dehy-
drated zeolites were contacted with gaseous SiF4 , preferably highly diluted with
nitrogen, at temperatures from ambient to about 200C and subsequently sub-
jected to an ammonium ion-exchange procedure in order to remove AlF2+ and
AlF2+ ions from the treated zeolite. Examples given in the patent specification
referred to the dealumination of Y zeolite, mordenite and ALPO4-5. The repor-
ted dealumination degrees, reflected by an increase in the Si/Al ratio from 7 to
about 10 for mordenite and 2.5 to 3.2 for Y, were rather moderate. Moreover, no
Dealumination Techniques for Zeolites 237

information was given on the degree of crystallinity of the dealuminated zeo-


lites. The smaller molecular size compared to that of SiCl4 could make SiF4 appli-
cable for the dealumination of zeolites with smaller pore openings. On the other
hand, this compound is extremely reactive and, therefore, difficult to handle.
That may be the reason why this dealumination method has not been further
studied.

4.3
Dealumination by (NH4)2[SiF6] Solutions

Early in the 1980s, Breck and Skeels developed a new method for the dealumi-
nation of medium- and large-pore zeolites. It was first described in a patent
[180] assigned to the Union Carbide Corp. (application filed in 1981) and then
presented at the 6th International Zeolite Conference in 1984 [181]. Their fun-
damental idea was to treat a zeolite slurried in water with an aqueous solution
of an agent which extracts aluminum from the framework, provides ligands for
the formation of a thermodynamically strongly favored, soluble aluminum com-
plex and serves as an extraneous source of silicon atoms filling up the frame-
work vacancies formed upon extraction of aluminum. Breck and Skeels realized
that only soluble hexafluorosilicate salts, especially the ammonium and lithium
salts, meet the requirements of such a process. The overall process of this de-
alumination process can be described by Eq. (6).
(NH4)+x [AlxSi yO(2x + 2y)]x + (NH4)2 [SiF6]
(6)
(NH4)+(x1) [Al(x1)Si(y+1)O(2x + 2y)](x1) + (NH4)3 [AIF6] .
It is believed that the process proceeds in two steps: (1) the removal of aluminum
from the framework and (2) the insertion of Si atoms in the lattice vacancies left
by aluminum release. In order to avoid too high concentrations of defect sites
leading to unstable products, the reaction rate of the first step should not exceed
that of the second one. Thus, the pH of the slurry must be considered as a cruci-
al parameter since it decisively controls the rate of aluminum extraction.
Typically, a 1 M solution of (NH4)2[SiF6] is added to an aqueous suspension of
a Y zeolite in amounts determined, according to the stoichiometry of reaction
Eq. (6), by the desired dealumination degree of the final product, and with a rate
of 0.005 moles of (NH4)2[SiF6] per minute and mole of aluminum in the zeolite
[181]. This addition rate is crucial in order to maintain the reaction pH at the
required value of about 6. Alternatively, the slurry can be buffered, by e.g.,
ammonium acetate, in order to provide control of the pH. To complete reac-
tion (6), the slurry is refluxed for some hours and the aluminum, removed from
the framework and complexed to [AlF6]3, is washed out from the product with
water.
The removal of aluminum from, and the isomorphous incorporation of sili-
con into, the framework of Y zeolite and mordenite was clearly evidenced in the
early paper [181] and again in a later publication [182]. It was also shown that
this new dealumination method requires ammonium or hydrogen forms of zeo-
238 H.K. Beyer

lites and that it is not applicable to sodium forms generally obtained as as-syn-
thesized products in zeolite synthesis. (In this respect the technique resembles
the hydrothermal dealumination procedures.) However, questions as to what
degree zeolite frameworks can be dealuminated and to what extent dealumina-
tion is associated with creation of lattice defects and lattice damages were only
stressed in later papers [183191]. It is also obvious that this technique is only
applicable to zeolites whose charge-compensating cations form soluble fluoro-
aluminate and hexafluorosilicate salts since otherwise such complex salts would
be deposited in the pores, thus inhibiting the dealumination reaction or the
removal of the reaction product by washing.
Garraln et al. [183] found that the final structural and compositional cha-
racteristics of NH4,Na-Y zeolite dealuminated by (NH4)2[SiF6] at pH = 6 as
described in [181] depended strongly on reaction time, temperature, interme-
diate treatment (washing and calcination) and the molar ratio of applied
(NH4)2[SiF6] to aluminum in the sample subjected to dealumination. Up to the
substitution of 30 Al per unit cell, corresponding to a dealumination degree
of about 50%, the structure of the final zeolitic product proved to be stable
and essentially free of extra-framework aluminum species. However, when
(NH4)2[SiF6] was applied in amounts adequate to cause, according to the stoi-
chiometry of Eq. (6), higher dealumination degrees, the crystal structure pro-
gressively collapsed. No further dealumination was observed when the
(NH4)2[SiF6] treatment was repeated after intermediate washing. On the other
hand, calcination of the product at 500C between two or more subsequent
treatments resulted in further dealumination accompanied, however, by a sub-
stantial loss of crystallinity.
These results are in line with the findings of Zi and Yi [184], who compared
the surface acidity and physical properties of Y zeolites dealuminated up to
Si/Al ratios of 6.8 by treatment with (NH4)2[SiF6] solution at pH 6 and
70C with those of a dealuminated sample prepared from Na-Y by H4EDTA
extraction (Si/Al = 4.2) and of a commercial ultrastabilized Y zeolite (Si/Al =
2.7). Isomorphous substitution of aluminum for silicon by treatment with
(NH4)2[SiF6] solution resulted in products with higher crystallinity and almost
free of aluminum debris. In the group of dealuminated zeolites thus prepared,
the temperature of lattice collapse was found to increase nearly linearly with
the decrease of the number of aluminum atoms in the unit cell. However,
IR spectra in the OH-stretching vibration range and chemical analysis gave
evidence that, especially at higher dealumination degrees (final Si/Al ratios
greater than about 5), up to 3040% of the vacancies left by aluminum extrac-
tion were not refilled by silicon during the secondary synthesis and remained as
structural defects. As expected, the surface acidity proved to be preferentially of
the Brnsted type and the acid strength increased as compared to the parent
Y zeolite.
Similar results were reported in a paper of Neuber et al. [185] dealing with the
spectroscopic and catalytic characterization of NH4-Y zeolite progressively de-
aluminated with (NH4)2[SiF6] according to [181]. Again, at higher degrees of de-
alumination, the process was found to be associated with the creation of defect
sites or vacancies which were large enough to admit pyridine even to the soda-
Dealumination Techniques for Zeolites 239

lite cages, and a substantial loss of crystallinity was observed at dealumination


degrees higher than 50%. Further, fluorine species could be detected in the zeo-
litic product after removal of about 50% of the framework aluminum. With pro-
gressive dealumination, higher activation energies were found to be neccessary
for the decomposition of the respective ammonium and pyridinium zeolites,
i.e., the strength of Brnsted acid sites associated with remaining framework
aluminum was increased.
In a comparative study of Y zeolites dealuminated by both hydrothermal
treatment and (NH4)2[SiF6] solution much lower Lewis and higher Brnsted acid
site concentrations were found at comparable aluminum contents in the iso-
morphously substituted products [102].
More recently, Matharu et al. [186] pointed again to the sensitivity of the de-
alumination procedure published in [180182] against reaction parameters
such as reaction temperature and time, rate of addition, pH, washing conditions,
etc. They also found that the maximum dealumination degree was limited to
about 50% and oxy-fluorinated aluminum species, trapped within the zeolite
cages, might have been retained which could not be removed by washing. Cor-
ma et al. [109] estimated the dealumination limit approachable without essenti-
al lattice destruction to be 25 Al per unit cell, corresponding again to a dealumi-
nation degree somewhat lower than 50%.
The development of strong acidity upon dealumination of Y zeolite according
to, and the retention of, not further identified fluorine species was also reported
by Lnyi and Lunsford [187]. Using a parent NH4,Na-Y zeolite with 54 Al per
unit cell, an abrupt decrease in crystallinity (about 50%) and an increase in the
amount of retained fluorine species were observed when the product approa-
ched a framework aluminum content of 26 Al per unit cell, i.e., again at a de-
alumination degree of about 50%. It is worth mentioning that these authors
found Y zeolites dealuminated by (NH4)2[SiF6] to be substantially resistant to
further dealumination upon steaming at 600C for 3 h. However, part of the
retained fluorine escaped as HF during the hydrothermal treatment.
In a series of Y zeolites dealuminated to different degrees (final average frame-
work Si/Al ratios about 36) with (NH4)2[SiF6], which retained above 80% of
their crystallinity, a much greater Si/Al ratio was found by XPS on the outer sur-
face of the crystallites [188]. The gradient is obviously due both to diffusion-
controlled dealumination and to a selective deposition of silica on the external
surface. In another paper [189], this dealumination process was found to be not
completely stoichiometric. About 16% of the amount of (NH4)2[SiF6] applied,
which was enough to approach a dealumination level of 26.6%, did not react
under conditions usually recommended. Non-reacted fluorosilicate, difficult to
remove by washing, may have remained in the sample and may have been res-
ponsible for the low resistance of the dealuminated products to hydrothermal
treatments.
Treatment with (NH4)2[SiF6] solutions was found to have no effect on the
porosity and total acidity of ZSM-5 zeolites, but it decreased the concentration
of aluminum atoms on the external surface of the crystallites and, hence, im-
proved the para-selectivity of ZSM-5 in the catalytic isomerization of m-xylene
[190]. Similarly, applying the technique described in [181] to the dealumination
240 H.K. Beyer

of NH4-mordenite, only a few extra-framework aluminum species correspond-


ing to a dealumination degree of maximum 12% were formed, but they could
not be extracted by washing [191]. It was suggested that aluminum species re-
leased from the framework remained blocked in the unidimensional channel
system and, therefore, dealumination was limited to the pore mouth. This result-
ed, due to shape-selective effects, in a large decrease of the amount of adsorbed
organic molecules and of the catalytic activity in m-xylene isomerization.
The dealumination of faujasite, mazzite and offretite with ammonia hexaflu-
orosilicate and the characterization of the products with various techniques
have been reported [192]. The maximum level of dealumination, which could be
achieved without loss of X-ray crystallinity, corresponded to 50% for faujasite
and 30% for mazzite and offretite. The dealumination capability was found to
depend on the texture of the crystals, which may have indicated that the process
was diffusion-controlled.
NH4-ZSM-5 was dealuminated up to Si/Al ratios of 100 in a multi-step process
consisting of (1) steaming at 350650C in presence of a not precisely defined
admixed phosphorus compound, (2) extraction with 0.2 M (NH4)2[SiF6] solu-
tion at 80C, (3) removal of Al and F ions by washing with water and (4) steam-
ing at 800C [193]. The X-ray crystallinity of the products was fully retained,
and, according to expectation, the total acidity decreased with increasing de-
alumination degree. However, it should be noted that the portion of stronger
acid sites is claimed to decrease with the progress in dealumination.
Corma et al. [194, 195] removed selectively extra-framework aluminum from
ultrastabilized Y zeolites by extraction with 0.4 M aqueous solutions of
(NH4)2[SiF6] at 95C. The complex salt has to be applied in amounts just suffi-
cient for the elimination of these species; application in excess leads to conco-
mitant removal of framework aluminum. In zeolites steamed under severe con-
ditions (at 700750C) part of the extra-framework aluminum, probably high-
ly condensed species, proved to be resistant to this treatment. The removal of
extra-framework aluminum resulted in characteristic changes in both the cat-
alytic activity of the samples in the isobutane/2-butene alkylation and the de-
activation rate.
In any case, this process, that proceeds according to e.g., Eqs. (7a) and (7b),
must result in silica as a reaction product which should be, if not removed in col-
loidal form, deposited in the zeolite pores or present as a separate phase.
AlOOH + (NH4)2[SiF6] SiO2 + NH4[AlF4] + NH4HF2 , (7a)

AlO+ + H2O + (NH4)2[SiF6] SiO2 + NH4[AlF4] + NH4HF2 + H+ . (7b)

4.4
Dealumination of Zeolites in Dry Mixtures with (NH4)2[SiF6]

The literature which has appeared in the field of zeolite dealumination with
(NH4)2[SiF6] since the pioneering report by Breck and Skeels [181] has dealt
exclusively with the optimization of the reaction conditions and with structural
and compositional consequences of this technique as well as with limitations in
Dealumination Techniques for Zeolites 241

its applicability and not with the modification of fundamental experimental fea-
tures. Only recently have Beyer et al. [196] reported on an essentially modified
procedure in which (NH4)2[SiF6] is applied, ground with the ammonium form of
the zeolite in crystalline state. It was found that, upon heating at 140190C, the
following reaction proceeds in such mixtures from left to right:

{AlO4/2}(NH4)+ + (NH4)2[SiF6] {SiO4/2} + NH4[AlF4] + NH4HF2 + NH3 (8)

where {AlO4/2} and {SiO4/2} refer to primary tetrahedral units of zeolite struc-
tures containing Si and Al, respectively, as T atoms. The escape of the gaseous
reaction product, ammonia, is obviously related to the thermodynamic driving
force of this reaction.
The progress of the dealumination reaction can be easily monitored by titra-
tion of the volatile reaction product, ammonia, evolved according to Eq. (8) and
depends on reaction temperature and time. Complete conversion of the applied
(NH4)2[SiF6] is normally achieved in 0.53 h. However, the reaction temperature
should not exceed 190C in order to avoid thermal decomposition of
(NH4)2[SiF6] and NH4[AlF4]. It is also recommended that the mixture be heated
in a stream of a dry inert gas up to the final reaction temperature at a slow rate
(e.g., 5 K/min) in order to remove most of the water adsorbed in the zeolitic com-
ponent before hydrolytic side reactions can start. Generally, at the reaction tem-
perature, most of the formed NH4HF2 will be stripped off by sublimation; the rest
can be extracted together with the reaction product NH4[AlF4] by washing with
water. Replacement of Al by Si in tetrahedral framework sites has been evidenced
by 27Al and 29Si MAS NMR spectroscopy and XRD (unit cell shrinkage).
In the case of Y zeolite, product inhibition was found to occur at a dealumi-
nation level corresponding to an incorporation of about 32 silicon per unit cell,
i.e., when each large cavity contained 4 [AlF4]. Supported by XRD observations
it was suggested that these anions are located at or near the cation sites SII co-
ordinatively bound to framework oxygen atoms of the six-membered rings
connecting sodalite cages and large cavities. Dealumination of L zeolite (with 8.7
Al per unit cell) was limited by product inhibition at a level of 3 [AlF4] per unit
cell corresponding to a dealumination degree of 35%. However, repeated deal-
umination using again 3 (NH4)2[SiF6] per unit cell resulted in the replacement of
a further three Al atoms, i.e., in a dealumination degree of about 70%, and total
dealumination was achieved after a third step [197]. It is worthwhile mentioning
that in the case of L zeolite the shrinkage of the unit cell was strongly aniso-
tropic; only the cell constant a decreased upon dealumination, while c showed
even a small but significant increase. This points to a selective substitution of
aluminum in T1 sites, i.e., in the 12-membered rings.
Similar dealumination behavior was observed for mordenite with 8 Al per
unit cell. Product inhibition appeared at a level of 4 [AlF4] per unit cell, i.e., at a
dealumination degree of 50%, and the unit cell contraction was found to be ani-
sotropic. A practically aluminum-free mordenite was obtained after repeating
once the dealumination procedure with 4 (NH4)2[SiF6] per unit cell. ZSM-5 zeo-
lite containing 4 Al per unit cell could be completely dealuminated in one step,
obviously due to the low initial aluminum content.
242 H.K. Beyer

The dry variant of the dealumination technique based on (NH4)2[SiF6] gives


dealuminated zeolites generally poor in lattice defects provided that the reaction
temperature is below the decomposition point of ammonium hexafluorosili-
cate. The dealumination agent is not applied in amounts exceeding the limits
due to product inhibition and the formed NH4HF2 is stripped off by sublimation
before the product is contacted with washing water. The modified procedure
seems to be superior to the original one reported in [180182] also as far as
applicability limitations are concerned, since it can be used, in contrast to the
original one (see [190, 191]), without essential restrictions for the dealumination
of ZSM-5 and mordenite as well.
In ferrierites, solid-state dealumination was found to be inhibited at rather
low dealumination degrees [198]. Applying crystalline (NH4)2[SiF6] in an amo-
unt equivalent to the ammonium content (1.9 per unit cell) of a completely ion-
exchanged ferrierite, only 0.7 silicon atoms per unit cell could be substituted for
framework Al. However, the zeolite prepared by solid-state synthesis from a
crystalline magadiite variety contained a larger amount of extra-framework
aluminum (0.52 Al per unit cell) in the form of cationic or oxidic species and lat-
tice defects (vacancies) revealed by IR bands typical of internal silanol groups
and of pyridine coordinatively bound to extra-framework aluminum. IR spec-
troscopic evidence was given that (NH4)2[SiF6] reacted with such aluminum spe-
cies according to Eq. (7a) and the over-all reaction:
2 AlO+ + (NH4)2[SiF6] SiO2 + 2 AlF3 + 2 NH4+ (9)
The framework vacancies were found to be filled up according to:
{(O1/2H)4} + (NH4)2[SiF6] {SiO4/2} + 4 HF + 2 NH4F (10)
where {SiO4/2} and {(O1/2H)4} refer to primary tetrahedral building units with Si
as T atom and vacancies of the hydroxyl nest type, respectively.
The limited degrees of dealumination obtained with ferrierite were due to
these favored side reactions and to the restriction of migration processes caus-
ed, in addition to the formation of NH4[AlF4] in the actual dealumination pro-
cess (see Eq. 8), by deposition of the products of the reactions (7a) and (9) in the
relatively narrow channels of ferrierite.

5
Alumination of Zeolites
The use of zeolites as ion exchangers generally requires high ion-exchange capa-
cities and, hence, high framework aluminum contents. Adsorption capacity and
selectivity also depends to a certain degree on the aluminum concentration of
the framework and may be favored by high framework aluminum levels. There-
fore, efforts have been made to find methods for the insertion of aluminum into
zeolite frameworks by secondary synthesis.
Dealumination Techniques for Zeolites 243

5.1
Alumination with Gaseous Aluminum Chloride

Very soon after the dealumination process based on the reaction with silicon
tetrachloride (see Sect. 4.1) was first reported [147], attempts were made to
reverse this reaction in order to increase the aluminum content of zeolitic fra-
meworks. High-siliceous zeolites of the ZSM-5 and ZSM-11 type were subjected
at 500600C to a stream of dry nitrogen loaded at higher temperatures
(180375C) with gaseous aluminum halides [199203]. 27Al MAS NMR, FTIR
and ammonia TPD techniques revealed that the content of both framework and
extra-framework aluminum was increased by this treatment, resulting in the
generation of both Brnsted and Lewis acidity. This behavior was suggested to
be due to the reversibility of reaction (5) [202]. However, Dessau and Kerr [199]
and Chang et al. [201] observed that, as far as the aluminum incorporation in
tetrahedral framework sites is concerned, internal hydroxyl groups associated
with lattice defects were involved in the process. Thus, they suggested the alumi-
nation phenomenon to be due to the insertion of Al in lattice vacancies.
For thermodynamic reasons it is not possible to consider the reverse of reac-
tion (4) as a pathway of framework alumination (which would formally result in
additional {AlO4/2} tetraeders charge-compensated by Al3+ cations) though that
has been done in some publications. Yashima et al. [204] found that aluminum
was not incorporated into HZSM-5 heat-treated at 960C prior to contacting
with AlCl3 at 350C. On the other hand, aluminum was inserted into the
framework of this zeolite upon contacting with gaseous AlCl3 at 650C. Silicon
released from the zeolite, obviously in the form of a volatile silicon compound,
was recovered by passing the effluent gas through 1 N NaOH and determined by
AAS. HZSM-5 pretreated at only 500C was found to react with AlCl3 already at
350C. On the basis of these observations, it was suggested [204] that at lower
temperatures, e.g. 350C, alumination proceeded through the reaction of
hydroxyl nests with AlCl3 , whereas at higher temperatures (650C) substitution
of framework silicon by aluminum, i.e., the reverse of reaction (5), occurred.
However, in a later paper [205], the same authors reported that alumination
levelled off within a certain reaction time while the amount of released silicon
increased steadily und surpassed that of the incorporated aluminum. Conse-
quently, it was suggested in line with preceding publications that the introduc-
tion of aluminum in four-coordinated framework sites proceeded exclusively
via insertion in lattice imperfections though no attempts were made to explain
the claimed presence of volatile silicon compound(s) in the effluent gas. The for-
mation of extra-framework aluminum species was attributed to the reaction of
AlCl3 with silanol hydroxyls on the external surface and/or the non-intact Si-O-
Si bonds formed from the SiOH groups on the external surface.
The reaction of AlCl3 with Y zeolites, disclosed in patents assigned to Mobil
Oil [206] and to Esso [207], was found to be basically a vapor-phase exchange
of the original zeolite cations with aluminum ions.Thus, at least a fraction of
the aluminum incorporated in ZSM-5 zeolites probably compensated as lattice
cations the negative framework charges created by the incorporation of alu-
minum into the framework.
244 H.K. Beyer

In contrast to findings reported in [201], framework alumination of HZSM-5


was claimed to generate very strong acid sites [205, 208] by a synergetic effect of
Lewis acid sites on Brnsted-type sites. However, since this claim was based only
on ammonia TPD results that gave no direct evidence for the chemical nature of
acid sites, it may be that the observed high-temperature TPD peak was exclus-
ively associated with typical Lewis acid sites.
Wu et al. [209] re-inserted aluminum in mordenite previously dealuminated
by acid leaching. The optimal temperature for the reaction with gaseous AlCl3
was found to be 600C. The amount of incorporated aluminum proved to be
proportional to the defect site concentration and exceeded the amount of releas-
ed silicon, depending on the reaction temperature, by a factor of 2580. Thus,
in the case of mordenite also, lattice vacancies are basically involved in the re-
insertion of tetrahedrally coordinated aluminum. Using GaCl3 and SbCl3 as
reactants, gallium and antimony could be similarly inserted into lattice defi-
encies created in mordenite by acid leaching [210].
Recently [211], a post-synthesis modification of zeolite beta consisting of
separate dealumination and titanation steps has been reported. First hydroxyl
nests were formed by removal of up to 90% of the aluminum by leaching with
oxalic or nitric acid, than up to 2 wt.% titanium was inserted into the lattice
vacancies without formation of TiO2 as a second phase by treatment with
gaseous TiCl4 at 500C.
The general conclusion to be drawn is that alumination based on the reaction
with gaseous aluminum chloride is restricted to zeolites containing framework
vacancies by synthesis and is restricted to the level of such lattice imperfections.

5.2
Alumination with Aqueous Fluoroaluminates

Insertion of aluminum into zeolites by aqueous fluoroaluminates was reported


by Chang et al. [201]. The applicability of potential reactants was restricted by
the low solubility of the fluoroaluminate salts. The reported alumination proce-
dure for silicalite comprised impregnation with a 0.02 M aqueous solution of
(NH4)3[AlF6] and drying the sample, containing about 0.1% AlF3, at 130C. The
pH of the (NH4)3[AlF6] solution had a decisive influence on the incorporation of
Al into the framework. In contrast to gaseous AlCl3, alumination with hexaflu-
oroaluminuminates proceeded also in the absence of defect sites. Thus, in this
case, direct substitution of framework silicon by aluminum, i.e., the reverse of
reaction (6), seemed to occur.

5.3
Alumination with Aluminate Solutions

Shihabi et al. [212] observed that the ion-exchange capacity and acid-catalytic
activity of high-silica ZSM-5 significantly increased when the zeolite was extruded
with alumina binders. These effects were attributed to the transfer of aluminum as
an aquospecies from the binder (g-Al2O3) to the zeolite during extrudation or
hydrothermal treatment and their incorporation into framework defects [213].
Dealumination Techniques for Zeolites 245

Zhang et al. [214] used sodium aluminate solutions as external aluminum


source. Y zeolites dealuminated by extraction with EDTA solution and by treat-
ment with SiCl4 were treated with 0.0251.0 M aluminate solutions adjusted to
pH 14 at 60C for 12 h. MAS NMR and FTIR spectrometry gave evidence that
after both dealumination pretreatments aluminum was re-inserted into the frame-
work. The re-alumination mechanism was thought to involve incorporation of
aluminum into lattice vacancies in the case of the EDTA-treated samples and
direct substitution of framework silicon in high-silica Y zeolite prepared by the
SiCl4 procedure. In contrast to this, only alumination of the crystal surface and
no significant changes in the bulk composition were found when high-silica Y
zeolite, prepared by dealumination with SiCl4, was treated with NaAlO2 solutions
under comparable conditions [118, 215, 216]. The products obtained proved to
be highly resistant to superheated steam, which was attributed to the formation
of a superficial layer of amorphous aluminosilicate upon treatment with sodium
aluminate solution.

5.4
Re-Insertion of Extra-Framework Aluminum

In 1980, Breck and Skeels [217] reported that hydroxyaluminum cations, formed
during ultrastabilization of Y zeolite through hydrolytic release of framework
aluminum, could be at least partly (20%) re-inserted into the framework vacan-
cies by titration with NaOH up to a pH of 1011.These results were later dis-
counted by Engelhardt and Lohse [218]; applying 29Si MAS NMR spectroscopy,
they could not find any re-insertion of aluminum in samples prepared exactly
according to the data given in [217]. The experimental conditions applied in
[217] and [218] are obviously too mild to achieve a detectable re-alumination,
since Liu et al. [219] succeeded in increasing the framework aluminum content
of ultrastabilized Y zeolite by about 48% upon treatment of 2 g of sample with
100 ml of a 0.25 M aqueous KOH solution at 80C for 24 h. This was controlled
by 29Si MAS NMR spectroscopy. The reported procedure was claimed to reverse
completely the process of ultrastabilization. Bezman [220] and Klinowski et al.
[221] confirmed conclusively the aluminum re-insertion reported in [219].
However, it was evidenced [220] that the process is associated with an amor-
phization of about 20% of the treated zeolite.A study by Hamdan et al. [222] also
indicated that aluminum atoms eliminated from the framework of Y zeolite by
hydrothermal treatment could be subsequently re-inserted into the framework
by treatment with KOH solutions at 80C. Crystallinity was found to be largely
retained in the process. However the Si,Al distribution proved to be significant-
ly different from that in the starting zeolite.
The effect of alkalinity on the re-alumination of Y zeolite previously deal-
uminated by SiCl4 was studied by Zhang et al. [214]. Treatment with NaOH solu-
tions at pH 12 and higher resulted in remarkable re-alumination. However, the
crystallinity decreased with increasing pH, reaching about 50% at pH 14.
Recently, Liu et al. [223, 224] investigated the effects of 0.0252.0 M KOH
solutions on the structure of Y zeolites previously dealuminated by ultrastabi-
lization and by extraction with EDTA and (NH4)2[SiF6]. As for ultrastabilized
246 H.K. Beyer

Y zeolite, the re-insertion mechanism involving the refilling of lattice vacanc-


ies was again confirmed. The samples dealuminated by exctraction with both
EDTA and (NH4)2[SiF6] were found to be practically free of extra-framework
aluminum. Nevertheless, a significant increase in the framework aluminum
concentration was observed upon treatment of both samples with KOH solu-
tions. Mainly based on the treatment effects on crystal morphology and con-
centration of silanol groups, this phenomenon was attributed to the dissolu-
tion of the outer silicon-enriched layer in the case of the EDTA-treated sample
and to the removal of framework silicon in the zeolite dealuminated with
(NH4)2[SiF6].
Partial re-insertion of aluminum into the framework of H-ZSM-5 dealumi-
nated by calcination at 800C was observed upon treatment with alkaline solu-
tions [225]. Similarly, non-framework aluminum species created by deep hydro-
thermal dealumination of H-ZSM-5 could be partially re-introduced into the
framework upon treatment (2 h, 77C) with 0.1 M NaOH solution [226]. How-
ever, no re-alumination was observed after dealumination under mild hydro-
thermal conditions when the Si/Al ratio of the dealuminated material was <25.
Lietz et al. [227] studied the effects of NaOH treatment on H-ZSM-5 zeolite
after both calcination and steaming of the ammonium form at 500C. The calcin-
ed sample with relatively high framework (4.9 Al/u.c.) and low non-framework
Al content (0.3 Al/u.c.) released upon treatment with a 0.08 M NaOH solution at
reflux predominantly silicon. This desilication resulted in a moderate decrease
in the framework Si/Al ratio and in a significant enrichment of non-framework
aluminum on the outer surface of the zeolite crystallites. In contrast, alkaline
treatment of the steamed sample with low framework (0.8 Al/u.c.) and high
non-framework Al content (4.2 Al/u.c.) caused a re-alumination of the frame-
work at the expense of extra-framework aluminum species. In line with these
findings, Sulikowski [29] reported on a significant decrease (from 2.24 to 2.03)
in the fractal dimension of steamed NH4 ,Na-Y zeolite upon treatment with
0.25 M KOH pointing to a reduction of the surface roughness which is due
to the formation of aluminate ions and their re-incorporation into the frame-
work.
Direct evidence in support of the re-alumination of Y zeolite by (NH4)[AlF4]
was obtained from the XRD patterns of NH4 ,Na-Y previously subjected to the
first step of the modified dealumination technique reported in [196] (see Eq. 8),
i.e., only to the heat treatment of the (NH4)2[SiF6]/zeolite mixture, but not to the
subsequent removal of reaction products by washing. According to expectation,
after completed dealumination only the reflections typical of the (partly de-
aluminated) faujasite phase, but not those of the dealuminating agent, were ob-
served.After heating the samples for a longer time at relatively low temperatures
(about 150C) in air, reflections typical of Na2[SiF6] appeared in the XRD pattern
[196]. Obviously, sodium ions migrate from sites in the sodalite cages into the
large cavities where they react with the occluded reaction products of the pre-
viously performed dealumination, i.e., [AlF4] and NH4HF2, under formation of
thermodynamically favored sodium hexafluorosilicate:

{SiO4/2} + [AlF4] + NH4HF2 + 2 Na+ {AlO4/2} + Na2[SiF6] + NH4+ + H+ (11)


Dealumination Techniques for Zeolites 247

Storage of zeolite Y, that had previously been subjected to the first step of the
dealumination procedure, over an aqueous solution of ammonia resulted even
at room temperature within some days in the formation of Na2[SiF6] in amounts
detectable by XRD [228]. Thus, disregarding the participation of sodium ions,
Eq. (11) can be regarded as the reverse of the process described by Eq. (8).
Recently [229], re-insertion of extra-framework aluminum formed in H-
ZSM-5 during preceding calcination at 600C was claimed to proceed upon
treatment with 2 M acidic solutions at 100C. Hydrochloric acid proved to be the
most effective re-alumination agent. The framework aluminum content was
found to increase at the expense of extra-framework species from 58% re-
maining after the thermal dealumination up to 81% of the original amount in
the parent material. However, this re-alumination process is obviously restricted
to ZSM-5, since the removal of aluminum from other zeolites by acid leaching
has been convincingly demonstrated in numerous reports (see Sect. 2.1). In-
terestingly, Kooyman et al. [20] reported that acid leaching of ZSM-5 with HCl
resulted, independently of temperature and time of the treatment, in about the
same, though rather insignificant, dealumination degree. Thus, it seems that in
the presence of acids some equilibrium between framework and extra-frame-
work aluminum is established in ZSM-5.

6
Desilication of Zeolites
In principle, desilication of zeolite frameworks must result in the same features,
i.e., in the same type of lattice vacancies, as their dealumination. Considering the
immense efforts made during the last three decades in the field of dealumina-
tion and re-alumination of zeolites it seems astonishing that for a long time no
attempts were made to manipulate directly the framework silicon content by
secondary syntheses. Only recently have some studies been published aimed at
the desilication of zeolites by leaching with alkaline solutions.
Removal of framework silicon from Y zeolites (Si/Al = 2.7) upon leaching
with alkaline solutions up to pH 12 at 80C was already discussed in 1988 [230]
and connected with the observed increase of the unit cell size (framework Al/Si
ratio). Later, Dessau et al. [231] reported that treatment of ZSM-5 with refluxing
0.5 M Na2CO3 solution resulted in partial dissolution of the sample with prefer-
ential removal of silica from the outer shell of the crystals and, hence, in alumi-
num zoning, with aluminum enriched at the exterior crystal surface.
Mao et al. [232, 233] extracted X, Y and ZSM-5 zeolites under similar condi-
tions, i.e., twice at 80C for 4 h with an aqueous 0.8 M Na2CO3 solution
(5g/150 ml). According to compositional values given in [233], 41 Si atoms per
unit cell could be removed from ZSM-5 with a starting Si/Al ratio of 15.7, while
the Si/Al ratio of X and Y zeolite could be reduced from 1.24 to 1.00 (21 Si/u.c.)
and from 2.39 to 1.30 (62 Si/u.c.), respectively. It was claimed that the original
structure, degree of crystallinity, surface area and size of micropores were all
essentially preserved. From microporosity measurements the authors conclud-
ed that a healing process occurred upon heat treatment resulting in a secondary
pore system.
248 H.K. Beyer

Removal of framework silicon from Y zeolite dealuminated with (NH4)3[AlF6]


and virtually free of extra-framework aluminum species was also suggested by
Liu et al. [224], who observed a gradual decrease of the framework Si/Al ratio
upon treatment with 0.25 M KOH solution at temperatures between 40 and
100C and the appearence of soluble silicate in the fluid. Moderate desilication
(about 10%) of H-ZSM-5 upon treatment with 0.08 N NaOH at reflux tempera-
ture was observed by Lietz et al. [227].
It is highly probable that lattice vacancies created by removal of silicon may
be filled up in the same way and under similar conditions as those remaining
after release of aluminum. Thus, recrystallization of the desilicated products to
particles with well-ordered crystal structure but traversed by nanopores is
obviously effected by water steam present as reaction product of the dehydroxy-
lation of hydroxyl nests. After the desilication process, the zeolite is in the
sodium (or potassium) form and this is known to be highly resistant towards
hydrothermal effects. Therefore, it is to be expected that steaming represents the
most effective method for the stabilization of desilicated zeolites avoiding the
risk of concurrent dealumination.

References
1. Scherzer J (1984) In: White TE, Dalla Beta RA, Derouane EG, Baker RTK (eds) Catalytic
materials: relationship between structure and reactivity. American Chemical Society,
Washington DC, p 157. ACS Symp Ser 248
2. Stach H, Lohse U, Thamm H, Schirmer W (1986) Zeolites 6:74
3. Karge HG, Weitkamp J (1986) Chem-Ing Techn 58:946
4. Sulikowski B (1996) Heterog Chem Rev 3:203
5. Barrer RM, Makki MB (1964) Can J Chem 42:1481
6. Zhdanov SP, Novikov BG (1966) Dokl Akad Nauk SSSR Ser Khim 16:1107
7. Belenkaya IM, Dubinin MM, Krishtofori II (1967) Isv Akad Nauk SSSR 1967:2164
8. Wolf F, John H (1973) Chem Techn 25: 736
9. Hernandez F, Ibarra R, Figueras F (1985) Acta Phys Chem (Szeged) 31:81
10. Kornatowski J, Rozwadowski M, Gutsze A, Wisniewski KE (1989) In: Weitkamp J, Karge
HG (eds) Zeolites as catalysts, sorbents and detergent builders applications and inno-
vations. Proceedings of an International Symposium, Wrzburg, September 48, 1988.
Elsevier, Amsterdam, p 567. Stud Surf Sci Catal 46
11. Bodart P, Nagy JB, Debras G, Gabelica Z, Jacobs PA (1986) J Phys Chem 90:5183
12. Karge HG, Dondur V (1990) J Phys Chem 94:765
13. van Niekerk MJ, Fletcher JCQ, OConnor CT (1992) J Catal 138:150
14. ODonovan AW, OConnor CT, Koch KR (1995) Microporous Mater 5:185
15. Maache M, Janin A, Lavalley JC, Joly JF, Benazzi E (1993) Zeolites 13:419
16. Lami EB, Fajula F, Anglerot D, Des Courieres T (1993) Microporous Mater 1:237
17. Briscoe NA, Casci JL, Daniels JA, Johnson DW, Shannon MD, Stewart A (1989) In: Jacobs
PA, van Santen RA (eds) Zeolites: facts, figures, future. Proceedings of the 8th Interna-
tional Zeolite Conference,Amsterdam, July 1014, 1989. Elsevier,Amsterdam, p 151. Stud
Surf Sci Catal 49A
18. Kraushaar B, van Hooff JHC (1988) Catal Lett 1:81
19. Kornatowski J, Baur WH, Pieper G, Rozwadowski M, Schmitt W, Cichowlas A (1992)
J Chem Soc Faraday Trans 88:275
20. Kooyman PJ, van der Waal P, van Bekkum H (1997) Zeolites 18:50
21. Barrer RM (1958) Proc Chem Soc 1958:99
22. Lee EFT, Rees LVC (1987) J Chem Soc Faraday Trans I 83:1531
Dealumination Techniques for Zeolites 249

23. Beyer HK, Belenykaja IM, Mishin IW, Borbly G (1984) In: Jacobs PA, Jaeger NI, Jir P,
Kazansky VB, Schulz-Ekloff G (eds) Structure and reactivity of modified zeolites. Pro-
ceedings of an International Conference, Prague, July 913, 1984. Elsevier, Amsterdam,
p 133. Stud Surf Sci Catal 18
24. Kerr GT (1968) J Phys Chem 72:2594
25. Kerr GT (1969) J Phys Chem 73:2780
26. Datka J, Kolodziejski W, Klinowski J, Sulikowski B (1993) Catal Lett 19:159
27. Ciembroniewicz A, Zlcivska-Jezierska J, Sulikowski B (1979) Polish J Chem 53:1325

28. Lohse U, Stach H, Thamm H, Schirmer W, Isirikjan AA, Regent NI, Dubinin MM (1980) Z
anorg allg Chem 460:179
29. Sulikowski B (1993) J Phys Chem 97:1420
30. Beaumont R, Barthomeuf D (1972) J Catal 26:218
31. Pickert PE (1972) US Patent 3 640 681, Union Carbide Corp
32. Wiecznikowski A, Rzepa B (1977) Rocz Chem 51:1955
33. Harvey G, Prins R, Crockett R, Roduner E (1996) J Chem Soc Faraday Trans 92:2027 34.
34. Apelian MR (1981) US Patent, Mobil Oil Corp
35. Apelian MR, Fung AS, Kennedy GJ, Degnan TF (1996) J Phys Chem 100:16577
36. Garwood WE, Chen NY, Lucki SJ (1976) US Patent 3 937 791, Mobil Oil Corp
37. Garwood WE, Chen NY, Bailar JC Jr (1976) Inorg Chem 15:1044
38. Garwood WE, Lucki SJ, Chen NY, Bailar JC Jr (1978) Inorg Chem 17:610
39. Liu X, Xu R (1989) J Chem Soc Chem Commun 1989:1837
40. Ward JW (1967) J Catal 9:225
41. Kerr GT (1969) J Phys Chem 73:2780
42. Beyer H, Papp J, Kall D (1975) Acta Chim (Budapest) 84:7
43. Karge HG (1975) Z phys Chem Neue Folge 95:241
44. Maesen TLM, Sulikowski B, van Bekkum H, Kouwenhoven HW, Klinowski J (1989) Appl
Catal 48:373
45. Breck DW, Skeels GW (1977) In: Bond GC, Wells PB, Tompkins FC (eds) Proceedings of
the 6th International Congress on Catalysis, London 1976, vol 2. The Chemical Society,
London, paper B4
46. Kerr GT (1982) J Catal 77:307
47. Skeels GW (1983) J Catal 79:246
48. Klinowski J, Thomas JM, Fyfe CA, Gobbi GC (1982) Nature 296:533
49. Stabenow J, Marosi L, Schwarzmann M (1976) Ger Patent 2510740, BASF AG
50. Fejes P, Kiricsi I, Hannus I, Kiss , Schbel Gy (1980) React Kin Catal Lett 14:481
51. Fejes P, Kiricsi I, Hannus I, Schbel Gy (1983) Magy Km Folyirat 89:264
52. Kiricsi I, Hannus I, Schbel Gy, Fejes P (1984) Magy Km Folyirat 90:529
53. Fejes P, Hannus I, Kiricsi I (1984) Zeolites 4:73
54. Fejes P, Kiricsi I, Tasi G, Hannus I, Bertti I, Szkely T (1989) Zeolites 9:392
55. Fejes P, Kiricsi I, Hannus I, Schbel Gy (1985) In: Drzaj B, Hocevar S, Pejovnik S (eds)
Zeolites: synthesis, structure, technology and application. Proceedings of an Internatio-
nal Symposium, Portoroz-Portorose, September 38, 1984. Elsevier, Amsterdam, p 263.
Stud Surf Sci Catal 24
56. Fejes P, Kiricsi I, Hannus I, Schbel Gy (1988) In: Kall D, Minachev KH (eds) Catalysis
on zeolites. Akadmiai Kiad, Budapest, p 205
57. Hannus I, Fonseca A, Kiricsi I, Nagy JB, Fejes P (1995) In: Beyer HK, Karge HG,
Kiricsi I, Nagy J (eds) Catalysis by microporous materials. Proceedings of ZEOCAT95,
Szombathely, Hungary, July 913, 1995. Elsevier, Amsterdam, p 155. Stud Surf Sci
Catal 94
58. Hannus I, Nagy JB, Kiricsi I, Fonseca A, Fernandez C, Fejes P (1995) Z phys Chem
(Munich) 189:229
59. Hannus I, Knya Z, Lentz P, Nagy JB, Kiricsi I (1999) In: Kiricsi I, Pl-Borbly G, Nagy JB,
Karge HG (eds) Porous materials in environmentally friendly processes. Proceedings of
the 1st International FEZA Conference, Eger, Hungary, September 14, 1999. Elsevier,
Amsterdam, p 245. Stud Surf Sci Catal 125
250 H.K. Beyer

60. Knya Z, Hannus I, Kiricsi I, Lentz P, Nagy JB (1999) Colloids Surfaces A: Physicochem
Eng Aspects 35:158
61. McDaniel CV, Maher PK (1968) In: Molecular sieves. Society of Chemical Industry,
London, p 186
62. Kerr GT (1967) J Phys Chem 71:4155
63. Kerr GT (1969) J Catal 15:200
64. Hansford RC (1967) US Patent 3 354 077, Union Oil Co
65. Ward JW (1970) J Catal 18:348
66. Maher PK, Hunter FD, Scherzer J (1971) In: Molecular sieve zeolites, vol 1. American
Chemical Society, Washington DC, p 266. Adv Chem Ser 101
67. Scherzer J, Bass JL (1973) J Catal 28:101
68. Gallezot P, Beaumont R, Barthomeuf D (1974) J Phys Chem 78:1550
69. Maxwell IE, van Erp WA, Hays GR, Couperus T, Huis R, Clague ADH (1982) J Chem Soc
Chem Commun 1982:523
70. Engelhardt G, Lohse U, Samoson A, Mgi M, Tarmak M, Lippmaa E (1982) Zeolites 2:59
71. von Ballmoos R (1981) The 18O-exchange method in zeolite chemistry: synthesis, cha-
racterization and dealumination of high silica zeolites. Dalle and Sauerlnder, Frank-
furt/Aarau, p 185
72. Lohse U, Mildebrath M (1981) Z anorg allg Chem 476:126
73. Zukal A, Patzelov V, Lohse U (1986) Zeolites 6:133
74. Patzelov V, Jger NI (1987) Zeolites 7:240
75. Sulikowski B, Ciembroniewicz A, Komorowska E (1986) Pol J Chem 60:255
76. Lynch J, Raatz F, Dufresne P (1987) Zeolites 7:333
77. Raatz F, Freund E, Marcilly C (1985) J Chem Soc Faraday Trans I 81:299
78. Chevreau T, Chambellan B, Lavalley JC, Catherine E, Marzin M, Janin A, Hmidy JF, Khab-
tou S (1990) Zeolites 10:226
79. Klinowski J, Fyfe CA, Gobbi GC (1985) J Chem Soc Faraday Trans I 81:3003
80. Boscek V, Mastikhin VM (1987) J Phys Chem 91:260
81. Rocha J, Klinowski J (1991) J Chem Soc Chem Commun 1991:1121
82. Alemany LB, Kirker GW (1986) J Am Chem Soc 108:6158
83. Ray GJ, Samoson A (1993) Zeolites 13:410
84. Kranich WL, Ma YH, Sand LB, Weiss AH, Zwiebel I (1970) In: Molecular sieve zeolites,
vol 1. American Chemical Society, Washington DC, p 502. Adv Chem Ser 101
85. Chen NY, Smith FA (1976) Inorg Chem 15:295
86. Scherzer J (1978) J Catalysis 54:285
87. Lohse U, Alsdorf E, Stach H (1978) Z anorg allg Chem 447:64
88. Boscek V, Patzelov V, Tvaruzkov Z, Freude D, Lohse U, Schirmer W, Stach H, Thamm H
(1980) J Catal 61:435
89. Fernandez C, Auroux A, Vedrine JC, Grosmangin J, Szabo G (1986) In: Murakami Y,
Iijima A, Ward JW (eds) Proceedings of the 7th International Zeolite Conference, Tokyo,
August 1722, 1986. Elsevier, Amsterdam, p 345. Stud Surf Sci Catal 28
90. Ferre G, Dufresne P, Marcilly C (1984) French Patent 84 13474, Institut Franais du Petrol
91. Fajula F, Lambret M, Figueras F (1989) In: Weitkamp J, Karge HG (eds) Zeolites as cata-
lysts, sorbents and detergent builders applications and innovations. Proceedings of an
International Symposium, Wrzburg, September 48, 1988. Elsevier, Amsterdam, p 61.
Stud Surf Sci Catal 46
92. Sulikowski B, Rakoczy J, Hamdan H, Klinowski J (1987) J Chem Soc Chem Commun
1987:1542
93. Debras G, Gourgue A, Nagy JB, De Clippeleir G (1986) Zeolites 6:241
94. Yoshida A, Inoue K, Adachi Y (1991) Zeolites 11:223
95. Gross Th, Lohse U, Engelhardt G, Richter K-H (1984) Zeolites 4:25
96. Andersen B, Fletcher JCQ, OConnor CT (1993) In: von Ballmoos R, Higgins JB, Treacy
MMJ (eds) Proceedings of the 9th International Zeolite Conference, Montreal 1992, vol
II. Butterworth-Heineman, Boston, p 591
97. Ray GJ, Meyers BL, Marshall CL (1987) Zeolites 7:307
Dealumination Techniques for Zeolites 251

98. de Carvalho AP,Wang QL, Giannetto G, Cardoso D, Brotas de Carvalho M, Ramoa Ribeiro
F, Nagy JB, El Hage-Ai Asswad J, Derouane EG, Guisnet M (1990) J Chim Phys 87:271
99. Wang QL, Giannetto G, Torrealba M, Perot G, Kappenstein C, Guisnet M (1991) J Catal
130:459
100. Wang QL, Giannetto G, Guisnet M (1991) J Catal 130:471
101. Lutz W, Lffler E, Fechtelkord M, Schreier E, Bertram R (1997) In: Chon H, Ihm S-K, Uh
YS (eds) Proceedings of the11th International Zeolite Conference, Seoul, Korea, August
1217, 1996. Elsevier, Amsterdam, p 439. Stud Surf Sci Catal 105A
102. Macedo A, Raatz F, Boulet R, Janin A, Lavalley JC (1988) In: Grobet PJ, Mortier JW,
Vansant EF, Schulz-Ekloff G (eds) Innovation in zeolite material science. Proceedings of
an International Symposium, Nieuwpoort, September 1317, 1987. Elsevier,Amsterdam,
p 375. Stud Surf Sci Catal 37
103. Auroux A, Ben Taarit Y (1987) Thermochim Acta 122:63
104. Hong Y, Gruver V, Fripiat JJ (1996) J Catal 161:766
105. Bamwenda GR, Zhao YX, Groten WA, Wojciechowski BW (1995) J Catal 157:209
106. Li X-W, Han M, Liu X-Y, Pei Z-F, She L-Q (1997) In: Chon H, Ihm S-K, Uh YS (eds) Pro-
ceedings of the 11th International Zeolite Conference, Seoul, Korea, August 1217, 1996.
Elsevier, Amsterdam, p 1157. Stud Surf Sci Catal 105B
107. Zhixiang C, Caian R, Guangming T (1997) In: Chon H, Ihm S-K, Uh YS (eds) Proceedings
of the11th International Zeolite Conference, Seoul, Korea, August 1217, 1996. Elsevier,
Amsterdam, p 2011. Stud Surf Sci Catal 105C
108. Corma A, Forns V, Pallota O, Cruz JM, Ayerbe A (1986) J Chem Soc Chem Commun
1986:333
109. Corma A, Forns V, Martinez A, Melo F, Pallota O (1988) In: Grobet PJ, Mortier JW,
Vansant EF, Schulz-Ekloff G (eds) Innovation in zeolite material science. Proceedings of
an International Symposium, Nieuwpoort, September 1317, 1987. Elsevier,Amsterdam,
p 495. Stud Surf Sci Catal 37
110. Lohse U, Lffler E, Hunger M, Stckner J, Patzelov V (1987) Zeolites 7:11
111. Patzelov V, Drahordov E, Tvaruzkov Z, Lohse U (1989) Zeolites 9:74
112. Grobet PJ, Geerts H, Tielen M, Martens JA, Jacobs PA (1989) In: Weitkamp J, Karge HG
(eds) Zeolites as catalysts, sorbents and detergent builders applications and innova-
tions. Proceedings of an International Symposium, Wrzburg, September 48, 1988.
Elsevier, Amsterdam, p 721. Stud Surf Sci Catal 46
113. Shannon RD, Gardner KH, Staley RH, Bergeret G, Gallezot P, Auroux A (1985) J Phys
Chem 89:4778
114. Aouali L, Jeanjean J, Dereigne A, Tougne P, Delafosse D (1988) Zeolites 8:517
115. Yoshida A, Adachi Y (1989) Zeolites 9:111
116. He M-Y, Shu X, Min E (1995) Proceedings of the International Symposium on Zeolites in
China, Nanking, p 2/242
117. Wan Y, Shu X (1995) Proceedings of the International Symposium on Zeolites in China,
Nanking, p 2/235
118. Lutz W, Zibrowius B, Lffler E (1994) In: Weitkamp J, Karge HG, Pfeifer H, Hlderich W
(eds) Zeolites and related microporous materials: state of the art 1994. Proceedings of the
10th International Zeolite Conference, Garmisch-Partenkirchen, Germany, July 1722,
1994. Elsevier, Amsterdam, p 1005. Stud Surf Sci Catal 84B
119. Lago RM, Haag WO, Mikovsky RJ, Olson DH, Hellring SD, Schmitt KD, Kerr GT (1986)
In: Murakami Y, Iijima A, Ward JW (eds) Proceedings of the 7th International Zeolite
Conference, Tokyo, August 1722, 1986. Elsevier, Amsterdam, p 677. Stud Surf Sci
Catal 28
120. Arsenova-Hrtel N (1998) PhD thesis, Freie Universitt Berlin; Ethylbenzol Dispropor-
tionierung als eine Testreaktion zur Charakterisierung von Zeolithkatalysatoren,
Mensch & Buch Verlag, Berlin, p 175
121. Khl GH (1973) In: Uytterhoeven JB (ed) Proceedings of the 3rd International Con-
ference on Molecular Sieves, Zurich, Switzerland, September 37, 1973. Leuven Univer-
sity Press, Leuven, p 227
252 H.K. Beyer

122. Sano T, Suzuki K, Shoji H, Ikai S, Okabe K, Murakami T, Shin S, Hagiwara H, Takaya H
(1987) Chem Lett 1987:1421
123. Campbell SM, Bibby DM, Coddington JM, Howe RF, Meinhold RH (1996) J Catal 161:
338
124. Motz JL, Heinichen H, Hlderich WF (1997) In: Chon H, Ihm S-K, Uh YS (eds) Proceed-
ings of the11th International Zeolite Conference, Seoul, Korea, August 1217, 1996.
Elsevier, Amsterdam, p 1053. Stud Surf Sci Catal 105B
125. Loeffler E, Lohse U, Peuker Ch, Oehlmann G, Kustov LM, Tholobenko VL, Kazansky VB
(1990) Zeolites 10:266
126. He M-Y, Zhou M, Shu X, Fu W (1995) In: Proceedings of the International Symposium on
Zeolites in China, Nanking, p 2/242
127. Cairon O, Sellem S, Potvin C, Manoli JM, Chevreau T (1995) In: Bonneviot L, Kaliaguine
S (eds) A refined tool for designing catalytic sites. Proceedings of the International Sym-
posium, Qubec, Canada, October 1520, 1995, p 513. Stud Surf Sci Catal 97
128. Musa M, TarinaV, Stoica AD, Ivanov E, Plostinaru D, Pop E, Pop Gr, Ganea R, Brjega R,
Musca G, Paukshtis EA (1987) Zeolites 7:427
129. Meyers BL, Fleisch TH, Ray GJ, Miller JT, Hall JB (1988) J Catal 110:82
130. Miller JT, Hopkins PD, Meyers BL, Ray GJ, Roginski RT, Zajac GW, Rosenbaum NH (1992)
J Catal 138:115
131. Kim J-H, Sugi Y, Matsuzaki T, Hanaoka T, Kubota Y, Tu X, Matsumoto M (1995) Micro-
porous Mater 5:113
132. Lee K-H, Ha B-H (1998) Microporous Mesoporous Mater 23:211
133. Fernandez C, Vedrine JC, Grosmangin J, Szabo G (1986) Zeolites 6:484
134. Carvalho AP, Brotas de Carvalho M, Rama Ribeiro F, Fernandez C, Nagy JB, Derouane
EG, Guisnet M (1993) Zeolites 13:462
135. Meier WM, Olson DH (1992) Atlas of zeolite structure types, 3rd edn. Butterworth-
Heinemann, London, p 158
136. Lillerud KP, Stcker M (1989) In: Weitkamp J, Karge HG (eds) Zeolites as catalysts, sor-
bents and detergent builders applications and innovations. Proceedings of an Interna-
tional Symposium, Wrzburg, September 48, 1988. Elsevier, Amsterdam, p 769. Stud
Surf Sci Catal 46
137. McQueen D, Fajula F, Dutartre R, Rees LVC, Schulz P (1994) In: Weitkamp J, Karge HG,
Pfeifer H, Hlderich W (eds) Zeolites and related microporous materials: state of the art
1994. Proceedings of the 10th International Zeolite Conference, Garmisch-Partenkir-
chen, Germany, July 1722, 1994. Elsevier, Amsterdam, p 1339. Stud Surf Sci Catal 84B
138. Massiani P, Chauvin B, Fajula F, Figueras F (1988) Appl Catal 42:105
139. McQueen D, Chiche BH, Fajula F, Auroux A, Guimon C, Fitoussi F, Schulz P (1996) J Catal
161:587
140. Dutartre R, de Mnorval LC, Di Renzo F, McQueen D, Fajula F, Schulz P (1996) Micro-
porous Mater 6:311
141. Chauvin B, Massiani P, Dutartre R, Figueras F, Fajula F, Des Courieres T (1990) Zeolites
10:174
142. Uytterhoeven JB, Christner LG, Hall WK (1965) J Phys Chem 69:2117
143. Kerr GT (1969) J Phys Chem 71:4155
144. Steel AT, Dooryhee E (1993) Zeolites 13:336
145. Tsolovski I, Minchev Ch, Senderov EE, Penchev V (1985) In: Drzaj B, Hocevar S, Pejovnik
S (eds) Zeolites: synthesis, structure, technology and application. Proceedings of an
International Symposium, Portoroz-Portorose, September 38, 1984. Elsevier, Amster-
dam, p 337. Stud Surf Sci Catal 24
146. Hong Y, Fripiat JJ (1995) Microporous Mater 4:323
147. Beyer HK, Belenykaja I (1980) In: Imelik B, Naccache C, Ben Taarit Y, Vedrine JC, Coudu-
rier G, Praleaud H (eds) Catalysis by zeolites. Proceedings of an International Sym-
posium, Ecully (Lyon), September 911, 1980. Elsevier, Amsterdam, p 203. Stud Surf Sci
Catal 5
148. Daubre M (1854) Compt Rend 34:135
Dealumination Techniques for Zeolites 253

149. Beyer HK, Belenykaja IM, Hange F, Tielen M, Grobet PJ, Jacobs PA (1985) J Chem Soc
Faraday Trans I 81:2889
150. Klinowski J, Thomas JM, Audier M, Vasudevan S (1981) J Chem Soc Chem Commun
1981:570
151. Martens JA, Geerts H, Grobet PJ, Jacobs PA (1990) J Chem Soc Chem Commun 1990: 1418
152. Anderson MW, Klinowski J (1986) J Chem Soc Faraday Trans I 82:1449
153. Klinowski J, Thomas JM, Fyfe CA, Gobbi GC, Hartmann JS (1983) Inorg Chem 22:63
154. Kubelkov L, Seidl V, Novkov J, Bednrov S, Jru P (1984) J Chem Soc Faraday Trans I
80:1367
155. Garraln G, Corma A, Forns V (1989) Zeolites 9:84
156. Ray GJ, Nerheim AG, Donehue JA (1988) Zeolites 8:458
157. Kubelkov L, Dudkov L, Bastl Z, Borbly G, Beyer HK (1987) J Chem Soc Faraday Trans
I 83:511
158. Sulikowski B, Klinowski J (1990) J Chem Soc Faraday Trans I 86:199
159. Shi ZC, Auroux A, Ben Taarit Y (1988) Can J Chem 66:1013
160. Auroux A, Shi ZC, Echoufi N, Ben Taarit Y (1989) In: Weitkamp J, Karge HG (eds) Zeolites
as catalysts, sorbents and detergent builders applications and innovations. Proceedings
of an International Symposium, Wrzburg, September 48, 1988. Elsevier, Amsterdam,
p 377. Stud Surf Sci Catal 46
161. Bosacek V, Freude D, Frhlich T, Pfeifer H, Schmiedel H (1982) J Coll Interf Sci 85:502
162. Stockenhuber M, Lercher JA (1995) Microporous Mater 3:457
163. Hey JM, Nock A, Rudham R, Appleyard IP, Haines AJG, Harris RK (1986) J Chem Soc
Faraday Trans I 82:2817
164. Siantar DP, Millman WS, Fripiat JJ (1995) Zeolites 15:556
165. Sulikowski B, Borbly G, Beyer HK, Karge HG, Mishin IW (1989) J Phys Chem 93:3240
166. Grobet PJ, Jacobs PA, Beyer HK (1986) Zeolites 6:47
167. Roland E (1989) In: Weitkamp J, Karge HG (eds) Zeolites as catalysts, sorbents and deter-
gent builders applications and innovations. Proceedings of an International Symposium,
Wrzburg, September 48, 1988. Elsevier, Amsterdam, p 645. Stud Surf Sci Catal 46
168. Lutz W, Gessner W, Bertram R, Pitsch I, Fricke R (1997) Microporous Mater 12:131
169. Flp V, Borbly G, Beyer HK, Ernst S, Weitkamp J (1989) J Chem Soc Faraday Trans I
85:2127
170. Klinowski J, Thomas JM, Anderson MW, Fyfe CA, Gobbi GC (1983) Zeolites 3:5
171. Namba S, Inaka A, Yashima T (1984) Chem Lett 1984:817
172. Namba S, Inaka A, Yashima T (1986) Zeolites 6:107
173. Thomas JM, Klinowski J, Anderson MW (1983) Chem Lett 1983:1555
174. Bartl B, Zibrowius B, Hoelderich W (1996) J Chem Soc Chem Commun 1996:1611
175. Weitkamp J, Sakuth M, Cong-yan C, Ernst S (1989) J Chem Soc Chem Commun 1989:
1908
176. Unverricht S, Hunger M, Ernst S, Karge HK,Weitkamp J (1994) In: Weitkamp J, Karge HG,
Pfeifer H, Hlderich W (eds) Zeolites and related microporous materials: state of the art
1994. Proceedings of the 10th International Zeolite Conference, Garmisch-Parten-
kirchen, Germany, July 1722, 1994. Elsevier, Amsterdam, p 37. Stud Surf Sci Catal 84A
177. Klinowski J, Anderson MW, Thomas JM (1983) J Chem Soc Chem Commun 1983:525
178. Fyfe CA, Gobbi GC, Kennedy GJ, De Schutter CT, Murphy WJ, Ozubko RS, Slack DA (1984)
Chem Lett 1984:163
179. Gortsema FP, Lok BMT (1983) Eur Pat Appl EP 100 544, Union Carbide Corp
180. Breck DW, Blass H, Skeels GW (1985) US Patent 4 503 023, Union Carbide Corp
181. Breck DW, Skeels GW (1985) In: Olson D, Bisio A (eds) Proceedings of the 6th Interna-
tional Zeolite Conference, Reno, USA, 1015 July, 1983. Butterworths, Guildford, Surrey,
UK, 1984, p 87
182. Skeels GM, Flanigen EM (1989) In: Jacobs PA, van Santen RA (eds) Zeolites: facts, figures,
future. Proceedings of the 8th International Zeolite Conference, Amsterdam, July 1014,
1989. Elsevier, Amsterdam, p 331. Stud Surf Sci Catal 49A
183. Garraln G, Forns V, Corma A (1988) Zeolites 8:268
254 H.K. Beyer

184. Zi G, Yi T (1988) Zeolites 2:232


185. Neuber M, Dondur V, Karge HG, Pacheco L, Ernst S, Weitkamp (1988) In: Guisnet M,
Barrault J, Bouchoule C, Duprez D, Montassier C, Prot G (eds) Heterogeneous catalysis
and fine chemicals. Proceedings of an International Symposium, Poitiers, March 1517,
1988. Elsevier, Amsterdam, p 461. J Stud Surf Sci Catal 41
186. Matharu AP, Gladden LF, Carr SW (1995) In: Beyer HK, Karge HG, Kiricsi I, Nagy JB (eds)
Catalysis by microporous materials. Proceedings of ZEOCAT95, Szombathely, Hungary,
July 913, 1995. Elsevier, Amsterdam, p 147. Stud Surf Sci Catal 94
187. Lnyi F, Lunsford JH (1992) J Catal 136:566
188. Wang QL, Torrealba M, Giannetto G, Guisnet M, Perot G, Cahoreau M, Caisso J (1990)
Zeolites 10:703
189. Wang QL, Giannetto G, Guisnet M (1990) Zeolites 10:301
190. Silva JM, Ribeiro MF, Rama Ribeiro F, Gnep NS, Guisnet M, Benazzi E (1995) React Kinet
Catal Lett 54:29
191. Silva JM, Ribeiro MF, Rama Ribeiro F, Benazzi E, Gnep NS, Guisnet M (1996) Zeolites
10:275
192. Chauvin B, Boulet M, Massiani P, Fajula F, Figueras F, Des Courires T (1990) J Catal
126:532
193. Limin H, Quanzhi L, Zhiyuan X, Ding G, Guoxing N, Zaiting L, Zhicheng S (1997) In: Chon
H, Ihm S-K, Uh YS (eds) Progress in zeolites and microporous materials. Proceedings of
the11th International Zeolite Conference, Seoul, Korea, August 1217, 1986. Elsevier,
Amsterdam, p 2003. Stud Surf Sci Catal 105C
194. Corma A, Forns V, Rey F (1990) Appl Catal 59:267
195. Corma A, Martinez A, Martinez C (1996) Appl Catal A: General 134:169
196. Beyer HK, Borbly-Pln G, Wu J (1994) In: Weitkamp J, Karge HG, Pfeifer H, Hlderich
W (eds) Zeolites and related microporous materials: state of the art 1994. Proceedings of
the 10th International Zeolite Conference, Garmisch-Partenkirchen, Germany, July
1722, 1994. Elsevier, Amsterdam, p 933. Stud Surf Sci Catal 84B
197. Beyer HK, Pl-Borbly G, unpublished results
198. Pl-Borbly G, Beyer HK (1999) In: Kiricsi I, Pl-Borbly G, Nagy JB, Karge HG (eds)
Porous materials in environmentally friendly processes. Proceedings of the 1st Inter-
national FEZA Conference, Eger, Hungary, September 14, 1999. Elsevier, Amsterdam,
p 383. Stud Surf Sci Catal 125
199. Dessau RM, Kerr GT (1984) Zeolites 4:315
200. Dessau RM, Chu CT, Kerr GT, Miale JN (1985) Eur Pat Appl EP 134 849, Mobil Oil Corp
201. Chang CD, Chu CTW, Miale JN, Bridger RF, Calvert RB (1984) J Am Chem Soc 106:8143
202. Anderson MW, Klinowski J, Xinsheng L (1984) J Chem Soc Chem Commun 1984:1596
203. Jacobs PA, Tielen M, Nagy JB, Debras G, Derouane EG, Gabelica Z (1984) In: Olson D, Bisio
A (eds) Proceedings of the 6th International Zeolite Conference, Reno, USA, 1015 July,
1983. Butterworths, Guildford, Surrey, UK, p 783
204. Yashima T, Yamagishi K, Namba S, Nakata S, Asaoka S (1987) In: Grobet PJ, Mortier JW,
Vansant EF, Schulz-Ekloff G (eds) Innovation in zeolite material science. Proceedings of
an International Symposium, Nieuwpoort, September 1317, 1987. Elsevier,Amsterdam,
p 175. Stud Surf Sci Catal 37
205. Yamagishi K, Namba S, Yashima T (1990) J Catal 121:47
206. Miale JN, Weisz PB (1967) US Patent 3 354 078, Mobil Oil Corp
207. Kearby KK (1972) US Patent 3 644 220, Esso Res Engine Co
208. Namba S, Yamagishi K, Yashima T (1987) Chem Lett 1987:1109
209. Wu P, Komatsu T, Yashima T (1995) J Phys Chem 99:10923
210. Wu P, Komatsu T, Yashima T, Nakata SI, Shouji H (1997) Microporous Mater 12:25
211. Krijnen S, Snchez P, Jakobs BTF, van Hooff JHC (1999) Microporous Mesoporous Mater
31:163
212. Shihabi DS, Garwood WE, Chu P, Miale JN, Lago RM, Chu CT-W, Chang CD (1985) J Catal
93:471
213. Chang CD, Hellring ST, Miale JN, Schmitt KD (1985) J Chem Soc Faraday Trans I 81:2215
Dealumination Techniques for Zeolites 255

214. Zhang Z, Liu X, Xu Y, Xu R (1991) Zeolites 11:232


215. Lutz W, Lffler E, Zibrowius B (1993) Zeolites 13:685
216. Lutz W, Lffler E, Fechtelkord M, Schreier E, Bertram R (1997) In: Chon H, Ihm S-K, Uh
YS (eds) Proceedings of the11th International Zeolite Conference, Seoul, Korea, August
1217, 1996. Elsevier, Amsterdam, p 439. Stud Surf Sci Catal 105A
217. Breck DW, Skeels GW (1980) In: Rees LVC (ed) Proceedings of the 5th International Con-
ference on Zeolites, June 26, 1980, Naples, Italy. Heyden, London, p 335
218. Engelhardt G, Lohse U (1984) J Catal 88:513
219. Liu X, Klinowski J, Thomas JM (1986) J Chem Soc Chem Commun 1986:582
220. Bezman RD (1987) J Chem Soc Chem Commun 19870:1562
221. Klinowski J, Hamdan H, Corma A, Fornes V, Hunger M, Freude D (1989) Catal Lett 3:363
222. Hamdan H, Sulikowski B, Klinowski J (1989) J Phys Chem 93:350
223. Liu D-S, Bao S-L, Xu Q-H (1997) In: Proceedings of the International Symposium in
China, Nanking, p 3/59
224. Liu D-S, Bao S-L, Xu Q-H (1997) Zeolites 18:162
225. Vlter J, Lietz G, Krschner U, Lffler E, Caro J (1988) Catal Today 3:407
226. Reschetilowski W, Schllner R, Freude D, Klinowski J (1989) Appl Catal 56:L15
227. Lietz G, Schnabel KH, Peuker Ch, Gross Th, Storek W, Vlter J (1994) J Catal 148:562
228. Borbly-Pln G, Beyer HK, unpublished results
229. Sano T, Uno Y, Wang ZB, Ahn C-H, Soga K (1999) Microporous Mesoporous Mater 31:89
230. Aouali L, Jeanjean J, Dereigne A, Tougne P, Delafosse D (1988) Zeolites 8:517
231. Dessau RM, Valyocsik EW, Goeke NH (1992) In: Higgins JB, von Ballmoos R, Treacy MM
(eds) Proceedings of the 9th International Zeolite Conference, Extended Abstracts and
Program, Butterworth-Heinemann, Stoneham MA, RP96
232. Mao RLV, Xiao S, Ramsaran A, Yao J (1994) J Mater Chem 4:605
233. Mao RLV, Ramsaran A, Xiao S, Yao J, Semmer V (1994) J Mater Chem 4:533

You might also like