Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Crystallization Mechanism of Zeolite Beta From

Download as pdf or txt
Download as pdf or txt
You are on page 1of 30

35

Applied Catalysis, 31 (1987) 35-64


Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

CRYSTALLIZATIONMECHANISM OF ZEOLITE BETA FROM

(TEA)20, Na20 and K20 CONTAINING


ALUMINOSILICATE GELS .

Joaquin Perez-Pariente*, Johan A. Martens and Peter A. Jacobs**

Laboratorium voor Oppervlaktechemie, K.U. Leuven,


Kardinaal Mercierlaan 92, B-3030 Leuven, Belgium.

*
On leave from : Instituto de Catalisis y Petroleoquimica, Madrid.

** author for all correspondence

(Received 10 July 1986, accepted 13 February 1987)

ABSTRACT
Zeolite BETA is prepared in hydrothermal conditions from tetraethyl orthosilicate,
sodium aluminate, tetraethylammonium hydroxide, sodium and potassium hydroxide and
water. Detailed chemical analysis of intermediate solid and liquid phases obtained
from a large number of gel compositions at different synthesis times as well as SEM
and XRD investigations of the solids allow to determine the crystallization
mechanism. After nucleation in solution zeolite BETA crystals grow by incorporating
TEA-aluminosilicate species from solution. The influence on this process of sodium,
potassium, tetraethylammonium, silica, alumina and alkalinity in the gel are
discussed.

INTRODUCTION
The crystallization of high-silica zeolites from aluminosilicate gels in

hydrothermal conditions is a very complex process. The kinetics of nucleation and


crystal growth of high-silica zeolitic phases are governed by several parameters,
the most important ones being the nature of the organic template, the silica
source, the alkali cations, the molar composition of the synthesis mixture as well
as its basicity and the synthesis temperature.
The best known system is the synthesis of ZSM-5 in presence of tetrapropylammonium
(TPA) ions. Data on the synthesis mechanism of other high-silica zeolites are
fragmentary and in many cases the information is restricted to the examples
described in the patents.
Zeolite BETA is such a scarcely studied zeolite. It is a Mobil proprietary
material first synthesized in 1967 by Nadlinger, Kerr and Rosinski 111. The silica
to alumina molar ratio of the as-synthesized framework is between 10 and 200. For

0166-9834/87/$03.50 0 1987 Elsevier Science Publishers B.V.


36

crystalization of zeolite BETA the synthesis mixture is usually prepared from a

mixture of sodium aluminate, silica, tetraethylammonium (TEA) hydroxide and water.

The zeolite adsorbs equal amounts of hexane and cyclohexane and has a pore volume

of 221 cm3/kg.
The crystal structure of zeolite BETA has not been published yet. The void
characteristics of zeolite BETA crystals obtained with the synthesis recipes
described in this paper have been determined with the help of the n-decane test
reaction for probing zeolite pores 121. The results obtained point to the presence
of 12 membered ring channels with cages (2-41, which are presumably the

intersections of the 12-membered ring pores 151. With respect to pore structure,
zeolite BETA is unique as no other high-silica zeolite with intersecting

12-membered ring channels is known at this time.


The presence of large intersecting pores in a highly silicious framework
garantees that zeolite BETA is a potential candidate for transformation into a

valuable acid catalyst. The alkylation of benzene with dodecene to

preparephenyldodecenesf61 and the formation of diphenylmethane out of benzene and


trioxane 171 illustrate its catalytic abilities. Dealuminated zeolite BETA can be
obtained by HCl treatment 181. Dealuminated zeolite BETA loaded with noble metal
behaves as a bifunctional zeolite catalyst which is able to dewax hydrocarbon
feedstocks mainly by isomerization of the waxy components 19-111. Pt-loaded H-BETA
obtained according to the presently described recipes shows excellent selectivities
for highly branched isomers in the conversion of normal paraffins 131 . This
dewaxing mode constitutes a considerable advantage over the processes using
medium-pore zeolites, such as ZSM-5, or small-pore zeolites, such as erionite. Over
such zeolites dewaxing occurs by hydracracking of the normal (selectofoning) or
normal and slightly branched (M-forming) paraffins into gaseous, and consequently
less valuable products.
The synthesis mechanism of medium-pore high-silica zeol.ites such as ZSM-5 has
been studied extensively. In principle two types of crystallization mechanisms have
been described : solid hydrogel phase transformation and liquid phase ion
transport. The synthesis mechanism of large-pore high-silica zeolites such as
zeolite BETA remains however unexplored.

The present work is devoted to the synthesis of zeolite BETA. From a thorough
characterisation of the solid and liquid intermediate phases obtained at different
times from a large number of different compositions of the synthesis gel, much
insight into the crystallization chemistry could be obtained.

EXPERIMENTAL
Aluminosilicate gels were prepared by stirring at room temperature appropriate
amounts of sodium aluminate (Hopkin and Williams - technical grade), sodium
hydroxide (Merck - 99% pure), water, tetraethylammonium (TEA) hydroxide (Fluka, 40%
37

aqueous, technical grade) and tetraethyl orthosilicate (Merck, 98% pure).


Once the hydrolysis of the organosilicon compound was finished, stirring in an open
vessel was continued for 24 h to evaporate the ethanol formed during hydrolysis.
The gels were then placed in stainless steel autoclaves of 150 cm3 and heated

statically at 373K. After selected time intervals, the autoclaves were immersed in
cold water and their contents were centrifugated at 14,000 r.p.m. The solids were
washed with deionized water till pH 8-9 and dried at 353 K in air.

TABLE 1. Molar gel compositions used for the synthesis of zeolite BETA (A1203 =l).

Notation Na20 (TEA)20 SiO2 NaCl


K2° H2°

A 1.6 0.36 5.0 20 240

1.5 0.54 7.5 30 360


B1*
1.5 0.54 7.5 30 360 3.0
Bl
1.8 0.54 7.5 30 360
B2* 2.5 0.49 6.8 30 360
B2** 7.8 0.11 1.5 30 360
B2
3.0 0.54 7.5 30 360
B3
4.5 0.54 7.5 30 360
B4
C 2.4 1.08 15.0 60 720

D 3.2 1.8 25.0 100 1200

The TEAOH technical grade solution was found to be contaminated with potassium
(7.5 g dms3). In all experiments the molar ratio K20/(TEA)20 in the starting gel
was therefore equal to 0.072. The molar ratio H20/ Si02 in all preparations was 12.
Table 1 summarizes the compositions of the synthesis mixtures used in the kinetic
study of the crystallization of zeolite BETA.
Chemical analysis of the solid phases was carried out after removal of water
and organic compounds by a calcination at 1173 K. The materials were dissolved by
successive treatments with HF, H2S04 and H20. Al, Na and K concentrations in the
resulting solutions were measured by atomic absorption. The amount of Si02 in the
solids was not determined separately but calculated by difference. The solids were
further analyzed by powder XRD for quantitative and qualitative phase
identification.
An automated powder diffractometer using Cu KU radiation and equipped with a Mac
Brawn position-sensitive detector was used. The XRD-pattern of the zeolite BETA
shown in Fig.1 exhibited the most .intense diffractions and was taken as the 100%
38

TABLE 2. XRD pattern of zeolite BETA

This study Patent literature

Interplanar Relative intensitya Interplanar Relative


intensityb

d-spacing (nm) (%) after Fig. 1 d-spacing (nm) ref.1

1.13 39 1.14 + 0.02 M-S


0.76 3 0.74 i 0.02 W
0.67 2 0.67 + 0.02 W
0.42 15 0.425 f. 0.01 W
0.40 100 0.397 ? 0.01 VS
0.30 12 0.30 f 0.01 W
0.21 6 0.22 f 0.01 W

a. measured as height of diffraction lines;

b. W = weak, M-S = medium strong, VS = very strong

degree 26

FIGURE. 1. XRD-pattern of the reference zeolite BETA sample (100% crystalline)


obtained from gel D after 7 days of synthesis.
39

crystallinity standard. For this sample the most important diffraction lines and
their intensities are collected in Table 2. The values of the d-spacings and the
relative intensities of the XRD-lines of the zeolite BETA sample obtained
correspond to the values used in the patent to claim this zeolite 111. All

preparations described were phase-pure. The degree of crystallinity of a particular

sample was determined comparing the area of the most intense diffraction peak
(degrees 20 = 22.37 and dhkl = 0.397) to that of the 100% crystalline sample.

FIGURE 2. Molar proportions of Si02, A1203 and Na20 in reaction mixtures for the
synthesis of zeolite BETA. ((TEA)20/ SiO2 = 0.25, K20/Si02 = 0.018;
synthesis time : 4 weeks; AM = amorphous, P = zeolite P, shaded area =
zeolite BETA)

The solids were also analysed for crystal size and morphology with an ISI - SS60
Scanning Electron Microscope. Thermoanalytical measurements were made on a Mettler
Thermobalance in the temperature range 298 - 973 K. The heating rate was 6 K mln-I,
the air or N2 flow was 3.4 dm3 h-l and the amount of sample was 0.030 g. Weight
40

loss curves (TG), differential weight loss curves (DTG) and DTA curves were
registrated simultaneously. Infra-red (1R) spectra of the lattice vibrations of
zeolite BETA were recorded on a PE-580B dispersive instrument using the KBr pellet
technique.

RESULTS AND DISCUSSION.


1. Compositional area for synthesis of zeolite BETA.
The stability area of zeolite BETA in the system Na20K20,(TEA)20,Si02,A1203and.
HZ0 was determined after 4 weeks of autoclaving. The different compositions of the
synthesis mixture giving rise to a zeolite BElA phase are depicted in two ternary
compositional diagrams (Figs.2 and 3).

FIGURE 3. Molar proportions of (TEA)20, NaZO and Al203 in reaction mixture for the
synthesis of zeolite BETA (K20/(TEA)20) = 0.071, Si02/A1203 = 100 - 10;
PHI = zeolite PHI, P = zeolite P, shaded area = zeolite BETA).
41

As can be seen in Fig.2, the crystallization of zeolite BETA at the values of


the (TEA)20/Si02 and K20/Si02 ratios mentioned occurs only for high Si02/A1203 and
SiO2/Na20 ratio's. In an attempt to increase the A1203 content, only amorphous
material was formed. At higher alkalinity, the most stable zeolite is zeolite P. In
absence of Al, no solid is obtained after four weeks. Thus, a silica polymorph

with the structure of zeolite BETA


can not be obtained by synthesis ;n the present
conditions (Fig.2).
On the other hand, a rather wide range of (TEA)20/Na20 ratios allows to
synthesize zeolite BETA (Fig.3). The A1203/(TEA)20 + Na20 ratio is, however,
restricted to a narrow range between 0.02 and 0.08. The use of high alkalinity
(Na20 + (TEA)20) is typical for the synthesis of this zeolite. Indeed, large
quantities of hydroxide ions are consumed during the hydrolysis of tretraethyl
orthosilicate and to neutralise the silicic acid thus formed. It should also be
noted that from the diagram of Fig.3 it does not result that BETA can be
synthesized in absence of inorganic alkali. Indeed, even when no extra sodium is
added, the K20/(TEA)20 ratio is still 0.071 as a consequence of the contaminated
TEA-OH solution used. The effect of the K20 concentration in the starting gel has
not been studied systematically. One experiment with Si02/A1203 = 30 and
(TEA)20/Si02 = 0.5 indicated that the use of a K20/(K20 + Na20) ratio of 0.35
resulted in the formation of merlinoite instead of zeolite BETA. A similar
behaviour has been reported for the synthesis of ZSM-5 1121. In K20 - rich gels
harmotome and feldspar are obtained instead of ZSM-5.
The crystallization mechanism and kinetics have then been studied in the
compositionally stable area as defined in Figs.2 and 3.

2. Chemical composition of solid and liquid intermediate phases.

2.1.Transport of silicon and aluminium during crystal growth


The chemical composition of the solid and the liquid phases changes
significantly during the crystallization of zeolite BETA.
Fig.4 illustrates how the silicon concentration in the liquid phase and the yield
of solids changes as a function of synthesis time for starting gels containing
various amounts of silica. The yield of solids is defined as the percentage of Si02
plus A1203 recovered in the solid phase. The crystallization curves can be divided
in three parts : an induction period followed by a period of exponential growth and
exhaustion of essential nutrients (Fig.4). During the induction period the silica
concentration in the liquid phase is very high for all samples and remains
unaltered. The yield of solids which can be recovered from such a system is low and
also does not change during this induction period. Intensive silicon transport from
the liquid to the solid phase occurs upon growth of the zeolite BETA crystals. The
decrease of the silicon concentration in the liquid phase occurs indeed in parallel
with the growth of the solid phase.
FIGURE. 4. Silica concentration in
the liquid phase (a) and yields of
solids (b) as a function of
synthesis time for gels with various
silica concentrations (notations :
11 I I I I I Table 1) The values in brackets refer
I b
to SiO,$ Al203 ratios of the gel.
(201,

(30) (

0 I I I I
0 5 10 15 20 25 23
TIME/DAYS

Fig.5 shows that the increase of the amount of recoverable solids is the
result of an increase of their crystallinity. This figure also indicates that the
amorphous solid which is rich in aluminium and is formed immediately from the
initial gel in synthesis conditions is hard to transform into zeolite BEIA. Indeed,

from aluminium-rich starting gels more amorphous solid is initially formed. If


large amounts of this solid are formed, never a highly crystalline BETA phase is

obtained.
The efficiency of the silicon consumption is relatively low since at the end of
the crystal growth high concentrations of silicon remain in solution. Between 30
and 60% of the available silicon is recovered as solid, even after very long
synthesis times (Fig.4a). The yield of solids formed during crystallization
increases considerably with decreasing Si02/A1203 ratio of the starting gel
(Fig.5).
From the molar compositions given in Table 1 the alumina concentration in the
initial mixtures can be calculated assuming that their composition is homogeneous.
When doing so, the alumina concentration of batches B1, B2, C and D is found to be
0.130, 0.130, 0.065 and 0.039 M respectively. The actual concentrations of alumina
43

as well as the Si02/Al203 molar ratio in solution for different synthesis times and
degrees of crystallinity of the solids are given in Table 3.

At any time the Al concentration in the liquids is two to three orders of magnitude
lower and the Si02/A1203 ratio's are two to three orders of magnitude higher than
the overall composition would predict. Most of the aluminium therefore has to be
located in the solid amorphous phase. Upon crystallization the alumina

concentration in solution decreases more rapidly than that of silica.

FIGURE 5. Correlation between the

yield of solids formed during


crystallization of zeolite BETA and

the degree of crystallinity of these


solids. Values in brackets refer to
different Si02/A1203 molar ratios of

the starting gel.

YIELD OF SOLIDS/ %

TABLE 3. Alumina (A1203) concentration and SiO,/ A1203 molar ratio in the liquid
phase at different synthesis times.

Gel Crystallization Si02/A1203a Degree of


A12O3
time (days) concentration ratio crystallinity of
(10-4M) solids (%)
in liquid phase

B1 0.6 12.4 3,000 0


2.0 8.1 12
28.0 1.3 17,000 82

B2 4.0 1.7 14,000 80


30.0 0.6 32,000 93

C 2.0 6.7 4,500 99


15.0 0.4 66,000 95

D 1.0 3.7 10,400 0


1.5 7.8 4,900 33
4.0 0.4 72,000 99

a. SiO2 in liquid phase determined as total Si02 minus Si02 recovered in the solids.
44

Indeed the Si02/A1203 ratio rises with crystallization time (Table 3). The very
high values of Si02/A1203 in solution are substantially above those commonly
reported for the synthesis of low-silica zeolites 113-191, but are in agreement
with the trace amounts of Al present in the mother liquor at the beginning of the
crystallization of ZSM-5 1201.

25- I

C
,
20-
0;

0" 61
B2 a
ix C A *
x nn
5- D

R2
‘D
‘61
I I I I
to-
0 20 40 60 00 1 0

degree of crystallinity/ %

FIGURE 6. Si02/A1203 molar ratio of the solid phase against its degree of
crystallinity for various compositions of the synthesis mixture
(notations : Table 1).

Fig.6 shows the evolution of the Si02/A1203 molar ratio of the solids obtained from
different compositions of the synthesis mixture as a function of the degree of
crystallinity. At the start of the synthesis when the solids are still X-ray
amorphous their Si02/A1203 molar ratio is typically between 1.5 and 3.0. Only the
addition of NaCl to the BI mixture (B1*) results in an enhanced silicon
incorporation in these amorphous solids. Since the liquid phase is very poor in

aluminium (Table 3) nearly all aluminium is present in the solids. This amount is
therefore determined by the aluminium content of the particular batch (Fig.4).
45

A very important observation is that all solids recovered during the induction
period do not contain a high concentration of nuclei of zeolite BETA, undetectable
by XRD,asthey do not show a specific IR-absorption at 525 or 575 cm-', which are
typical absorptions for zeolite BETA (Fig.7). X-ray amorphous ZSM-5 nuclei are
easily detected by the same procedure 1211. This indicates that during the
induction period BETA nuclei are either present in low concentration and not
detectable by IR or remain in solution given their lower density (compared to
ZSM-5) or their smaller size. As will be shown later fully crystalline BETA
consists of small crystals which makes the latter assumption the more probable one.

/-

$$\/i

:_

10 a
c-
L
3
%
I
a
2/
3
I

4
I~
0 2 IO
wavenumber / cm-l

FIGURE 7. IR spectrum of the solids obtained from sample B2 at a,12 h synthesis (0%
crystallinity), b, 24 h synthesis (0% crystallinity) and c, 30 days
synthesis (95% crystallinity).

The Al content of a particular synthesis mixture has no influence on the


overall SiO,/Al,O, ratio of partially crystalline solids, at least not up to medium
levels of crystallinity (Fig.6). Solids obtained from mixtures A, C and D with
Si02/A1Z03 ratio's of '20,60 and 100, respectively and otherwise identical K20/Si02,
(TEA)20/Si02 and H20/Si02 ratios (Table l), at 30% crystallinity all show a

Si02/A1203 molar ratio between 5.5 and 6.5 (Fig.6). Upon further crystallization
the solids become then more and more siliceous. Since the solids are composed of

amorphous material and zeolite BETA crystals, it is difficult to correlate the

Si02/A1203 ratio in the crystals to that of the starting gel. To do so, samples
with at least 95% crystallinity were selected and the Al/Si + Al fraction of these
46

solids was plotted against its value in the starting mixture (Fig.8).

I
0 0.05 0.10 0.15

AI/AI+Si in gel

FIGURE 8. Al fraction in a solid phase with at least 05% crystallinity against the
Al fraction in the starting mixture for zeolite BETA (0) (notations :
Table 1) and ZSM-5 after ref.22(o) ref.23 (A) and ref.24 (m). Dashed
lines represent hypothetical compositions of solids with a given
efficiency for silicon incorporation in the zeolite phase assuming that
all Al is incorporated. The valties in brackets refer to these
efficiencies.

Literature data on ZSM-5 synthesis indicate that the composition of the zeolite
is determined by that of the gel (Fig.8, points on the diagonal line), or in the
worse case the correlation between zeolite and gel composition is G.75. Changes in
the gel composition for zeolite BETA synthesis result in parallel but not in
proportional changes in zeolite composition. For BETA the correlation coefficient
between both compositions changes from 0.54 (run A) to 0.22 (run D). Since all
aluminium is incorporated in the zeolite (Table 3) this reflects again the
inefficiency with which silicon is incorporated into zeolite BETA. This occurs much
more readily during crystallizations from Al-rich gels. However, there exists an
upper limit for this behaviour since for the synthesis run A, never a
crystalllnity higher than 75% is reached (Fig.5).
When no aluminium is added to the starting gel D, also no solid phase is
recovered even after 14 days. This result is in agreement with the patent
literature iI/ in which only Si02/A1203 ratio's lower than 200 are claimed for
zeolite BETA, In this respect zeolite BETA differs from other high-silica zeolites
such as ZSM-5 and ZSM-11 for which Al-free counterparts are known, denoted with
silicalite-1 125) and silicalite-2 1261, respectively. For low-silica zeolites,
increasing the Al content in the gel increases the Al content in the zeolite too.
This was already recognized in the early work on synthesis of faujasites 1271.
47

2.2 Location of tetraethylammonium entities.

The incorporation of TEA as ion in the solids can be monitored indirectly via
the molar ratio Al203/(Na20 + K20). This ratio is plotted in Fig.9 against the
degree of crystallinity of the solids.

0.5 D I I I I I
0 20 40 60 80 100

degree of crystallinity /oh

FIGURE 9. Molar ratio of A1203/(Na20 + K20) of the solid phase against its degree
of crystallinity (notations : Table 1).

A1203/(Na20 + K20) ratio's exceeding unity require the incorporation of TEA+

cations into the zeolitic phase for reasons of charge balance. Fig.9 shows that the

incorporation of TEA' occurs certainly above 70% crystallinity, especially for the
most siliceous preparations (C and D). At lower degrees of crystallinity (ZO-60%)
this ratio is within experimental error equal to one. From the data presented in
Fig. 9 it can not be excluded, however, that TEA' is also incorporated in
materials, with a crystallinity below 70 %. The amorphous part occludes alkali
metal ions (at 0 % crystallinity the ratio A1203/(Na20 + K20) is lower than unity)
and the solids of intermediate degree of crystallinity are composed of amorphous
material next to zeolite BETA crystals.
Direct information regarding the incorporation of TEA' or TEA-OH is obtained
from thermoanalytic experiments. The TG - DTG - DTA pattern of a sample of zeolite
BETA with 100 % crystallinity is shown in Fig.10.
Four distinct zones of weight loss can be distinguished in the following
temperature zones : I : 298 to 493 K; II: 493 to 623 K; III : 623 to 773 K and IV :
773 to 973 K. The first step is slightly endothermic and results from water
desorption. The other three steps are exothermic and are related to the oxydation
of organic material. Parker et al. 1241 reported a weight loss around 623 K in the
decomposition of TEA-silicalite under helium flow and two maxima in the rate of
weight loss, around 623 K and 753 K for TEA in ZSM-5. The low temperature peak was
ascribed to the decomposition of TEA-OH occluded in the zeolite, while the
high-temperature maximum corresponded to the pyrolysis of TEA' cations. This
48

interpretation can be used to assign the second and third weight loss zone observed
for zeolite BETA (Fig.10) to occluded TEA-OH and TEA+ cations, respectively. The
exothermic weight loss at 863 K (IV), has to be caused by oxydative decomposition
of organic materials occluded in the zeolite. Support for the latter assignment was
obtained from another thermoanalytic experiment (Fig.11). When the decomposition ot
the organics in the zeolite is carried out under nitrogen instead of air, the
weight loss above 773 K is not observed, while the other two exotherms assigned to
TEA-OH and TEA+ combustion are transformed into endotherms corresponding to

pyrolysis of the same species. When this sample is subsequently cooled down from

TEMPERATURE/K

FIGURt 10. TG - DTG - DTA in air of 100% crystalline zeolite BETA prepared form

synthesis gel 0.

973 K to 873 K and air is introduced subsequently a very small exotherm is


detected, corresponding to a weight loss of 0.4 % . The resulting total weight loss

is then the same as in air (Table 4.)


For tetramethylammonium cations oxidative decomposition temperatures above

773 K have also been reported for zeolites OMEGA and offretite 129-301, and also
for zeolite ELI-1 prepared in presence of dialkylammonium cations 1311. Calcination
overnight of the present zeolite BETA samples at 823 K showed a drop in the degree
of crystallinity.
The total N-content of sample BE, at two different levels of crystallinity was

determined with the Kjeldahl method. Only when weight loss II is assigned to

TEA-OH removal and III + IV correspond to TEA+ decomposition, the total N-content
is in agreement with the weight losses observed (Table 5).
49

EXO

TIME /min.

FIGURE 11. TG - DTG - DTA pattern in nitrogen of zeolite BETA with 95%
crystallinity and prepared from synthesis mixture 62. Air is introduced
after cooling from 973 K to 873 K.

TABEL 4. Weight loss (%) of a zeolite BETA sample (95% crystallinity) prepared from
batch 62 and heated under nitrogen and air flow (Fig.11).

weight loss (%)

temperature zone I II III IV lva 11+111+1v

air 4.9 3.3 6.7 5.2 15.2

nitrogen 4.0 4.9 10.1 0.4 15.4

a. cooling from 973 K to 873 K and subsequent introduction of air.

When the sample is heated in air, some products of the oxydative decomposition of
TEA' might be strongly bound to the Al-sites, perhaps at not easily accessible
locations in the structure.
50

The results of the thermogravimetric analysis of an amorphous solid in air are


shown in Fig.12. Exotherms III and IV are totally absent while exotherm II is
virtually absent. Only a continuous weight loss between room temperature and 623 K
corresponding to the dehydration/dehydroxylation of this solid and to some extent
to decomposition of occluded TEA-OH is registered. This indicates that almost no
TEA is assGciated with the gel particles but is exclusively present in solution
before being incorporated in the zeolite BETA crystals.

25-

I I I I I I I I I
?100 400 500 600 700 800 900

TEMPERATURE/K

FIGURE 12. TG-DTG-DTA pattern in air of gel particles (0% crystallinity) obtained
from preparation 62 after 12 h synthesis.

Table 5. N-content (%) of zeolite BETA sample B2

crystallinity/% 40 95

N-content/%

from thermobalance (weight loss) 1.1 1.6

from Kjeldahl [N-content) 1.0 1.5


51

For two different compositions of the synthesis mixture, B2 and D, the weight
losses I, II and III + IV were determined for different degrees of crystallinity
(Figs.13 and 14).

degree of crystallinity /%

FIGURE 13. Thermogravimetric weight loss of the solids obtained from synthesis
mixture B2 as a function of the degree of crystallinity. The Roman
figures correspond to the zones defined in Fig. 10.

In many aspects the weight loss pattern of sample D with a Si0,/Al,03 ratio of 100
(Fig.14) is similar to that of B2 with a Si02/A1203 ratio of 30 (Fig.13) :
(i) absence of significant amounts of organics in the amorphous solids,
(ii) the amount of water released from the solids decreasing with increasing
degree of crystallinity,
(iii) weight losses corresponding to the decomposition of TEA+ cations (III + IV)
increasing with increasing degree of crystallinity.

Near 100% crystallinity, the relative amounts of water and TEA-species strongly
differ for these two samples. Sample D contains less water and more occluded TEA-OH
52

compared to 62. This higher hydrophobicity of sample


D is in line with its higher
silicon content (Fig.6). The weight loss above 623 K for samples 62
and D and their
degree of crystallinity are related via a straight line through the origin (Figs.13

and 14). TEAT incorporation clearly accompanies crystal growth and the amount of
TEA+ incorporated per unit weight of zeolite remains constant during
crystalTlzatiOn, irrespective of the Si02/A12D3 ratio of the starting gel.

127

ll-

O-

9-

X 8-
\ A
m
7-
:
6-
z
._01
5-
:

u 4-
I-

degree of crystallinity/%

FIGURE 14. Thermogravimetric weight losses of the solids obtained from synthesis
mixture D as a function of the degree of crystallinity. (Roman figures
represent the temperature zones of Fig.10).

From the total weight loss above 623 K (zone III + IV) corresponding to TEA+
loss and the chemical composition of the solids, the molar ratio N/Al can be

derived. The effect of the Al content of the zeolite on the fraction of the

negative charge compensated by TEA+ is shown in Fig.15.


It seems that zeolite BETA incorporates more TEA+ per Al when it iS more
siliceous. Zeolite BETA has a pore volume determined by cyclohexane adsorption of

221 cm3/kg (11. Using appropriate radii fur TEA+ and TEA-OH 132) from the

thermoanalytic data an apparent pore volume of 201 cm3/kg can be calculated for
53

sample D. Of this pore volume 68% is occupied by TEA+ and filled up with occluded
TEA-OH. This constitutes a strong argument for a pore filling function of the
organic template during synthesis. A similar mechanism has been advanced for ZSM-5
synthesis 1331.

OL
0 0.05
I I
0.10
I
0.15
I
0.20

Al /Al + Si

FIGURE 15. Molar ratio of TEA+/Al against the Al fraction in zeolite BETA at 80 to
100% crystallinity, (notations of Table 1)

3. A synthesis mechanism for zeolite BETA

The data presented in previous section allow to characterize the solid and
liquid phases in the synthesis of zeolite BETA in the following terms :
- a typical gel in synthesis conditions is transformed initially into a solid which
contains most of the aluminium (Table 3 and Fig.6) and a minor part of the
silicon; most of the silicon and all TEA species remain in solution
(Figs.4a and 12); from gels richer in aluminium more solids are formed (Fig.4b).
- during the nucleation or induction period the amount of solids does not change
(Fig.4b), the silica concentration in solution is not affected (Fig.4a), and no
typical framework vibrations are detectable in the solids by IR (Fig.7). The
amorphous solids contain occluded NaOH and KOH (Fig.9) but only traces of TEA-OH
(Fig.12). This amorphous aluminarich sodium (potassium) aluminosilicate is not
transformed as such into zeolite BETA since conditions yielding high amounts of
this solid are not suitable for the formation of highly crystalline BETA
(Fig.5). This suggests that during this period only small nuclei are formed
which remain in solution even when subjected to considerable centrifugal forces
(14,000 rpm).
54

- during the exponential growth curve following the induction period, the following
phenomena become apparent :

1. the decrease of the silicon concentration in the liquid phase occurs in


parallel with the growth of the solid phase; this requires Si transport from
the liquid to the solid phase (Fig.4,a,b); the crystallization efficiency on
a silica basis remains however unusually low (Fig.8).
ii. the increase in the yield of recoverable solids is the result of
crystallization (Fig.5).
iii. the initially present Al-rich amorphous sodium (potassium) aluminosilicate
phase is very hard to transform or incorporate into BETA crystals (Fig.5).
iv. the Si02/A1203 ratio of the solid phase increases during crystallization
(Fig.6); the occurrence of extremely high values of this ratio in solution
SUggeStS that aluminium is fixed in the zeolite lattice as soon as it is
available in solution after dissolution from the amorphous solid.
V. the efficiency for silicon uptake from solution is related to the initial gel
composition : the more Al in the gel, the more efficient is the silicon
incorporation in the crystals. However, there exists an upper limit for this
concentration, above which no fully crystalline materials can be obtained.
vi. the amount of TEA+ incorporated is linearly related to the degree of
crystallinity , irrespective of the gel composition (Figs.13 and 14) :
therefore the amount of TEA+ incorporated per unit weight of zeolite remains

constant; the SiO,/A1,03 composition of the starting gel determines TEA-OH

occlusion : for a zeolite rich in silica the sum of TEA+ and TEA-OH is close

to the pore volume of the structure; in a more hydrophylic zeolite obtained


from an Al-rich gel the extra Al incorporated is neutralized by alkali
cations; since the pore volume remains constant the amount of occluded TEA-OH

will decrease ; the constant amount of TEA+ taken up per unit weight of
zeolite suggests that Al associated with it is homogeneously distributed
throughout the crystal and no Al-gradients are generated; it is our
experience, however, that after reaching the maximum degree of crystallinity
the extra-amount of Al incorporated in association with inorganic alkali
decreases with time what may generate a decrease in Al concentration from the
crystal center to the outer rim.
Based on these observations the mechanism for crystal growth is schematically
presented in Fig.16. Three main steps are involved :

- dissolution of aluminosilicate gel particles.


- formation of high-silica aluminosilicate species in solution in a reaction
between dissolving aluminium and silicon with TEA-silicate species, abundantly
present in the liquid phase.
- condensation of the adequate precursors into the zeolite BETA framework.
55

Since the Al concentration in the liquid phase continuously decreases with time
(Table 3), an equilibrium between the precursors Ai in the liquid phase and the

amorphous gel particles is not established. The reaction of the soluble precursors
to form the zeolite can therefore not be the rate-determining step of the
crystallization process. It seems more probable that the rate of dissolution of the
amorphous aluminosilicate (kI) or the formation of adequate precursors (kp) is rate
controlling.

TEA-Silicate

h Ki
TEA-Silicate+ Al( + Na+ AI + Ay+
k:! I

I
k3

Zeolite BETA

FIGURE 16. Synthesis mechanism of zeolite BETA (Ai : high-silica aluminosilicate


species in solution).

The observation that no aluminium-free zeolite BETA can be obtained


points to the necessity of at least one Al atom in the construction of adequate
precursor species. The present synthesis of zeolite BETA is clearly not a
solid-state transformation of amorphous gel particles. The amorphous solids indeed
do not contain the template and their Si0,/A1203 ratio is different from that of
the zeolite. Further, the crystal growth involves clearly transportation through
the liquid phase of all elements of the crystals : Si, Al, Na, K and TEA.

4. Aspects of crystallization kinetics.

4.1. Influence of alkalinity.


The influence of the alkalinity of the starting mixture on the crystallization
kinetics has been studied by varying the Na20/Si02 molar ratio between 0.05 and
0.15 for otherwise identical compositions, based on silica (batches BI, B2, B3 and
B4-Table 1). The crystallinity of these samples as a function of the synthesis time
is given in Fig.17.
A maximum rate of crystallization is observed for a Na20/Si02 molar ratio of 0.06
(sample B2). Further increase of alkalinity leads to an increase of the induction
period and a reduction of the crystallization rate. The high hydroxyl
concentration in batch B3 causes zeolite BETA to dissolve after 15 days. At still
higher alkalinity zeolite P crystallizes together with zeolite BETA resulting in a

deviating behaviour of the curve of sample B4.


56

time /days

FIGURE 17. Degree of crystallinity of samples BI,B2, BS and B4 as a function of

time.

The increase of the alkalinity has been realized by addition of NaOH. Since this
increases the Na + concentration as well, the effect of Naf was checked in a
separate experiment by adding NaCl (batch 6: ). As can be seen from Fig.18 Na+ has
only minor influence on the nucleation rate as well as on the crystal growth rate.
So, the strong effects of NaOH addition displayed in Fig.17 are due to the addition
of OH-.

IO 15 20 25

time /days

FIGURE 18. Effect of addition of NaCl to the synthesis mixture BI on the degree of
crystallinity as a function of synthesis time.
Optimum hydroxyl concentrations have also been reported for the synthesis of
ZSM-5 122,34-371 and for low-silica zeolites such as A 138-401, faujasite 1411 and
mordenite 1421.
Mechanistically, an increase in alkalinity increases the rate of gel dissolution.
If this is the rate-determining step the overall crystallization process should
accelerate as is observed in batch B2 compared to B1. However there exists an
optimum OH- concentration since the precursors are involved in equilibria with
other species (Fig.16) and Kl, K2, . . . are probably pH-dependent functions. Very
alkaline media may affect the concentration of precursors leading to a slower
nucleation and crystal growth (Fig.17).
The kinetic effects caused by alkalinity changes are reflected in the crystal
size and morphology to the zeolite BETA samples. The crystals are spherically
shaped and only for the largest crystals a cubic morphology appears (Fig.19).

FIGURE 19. SEM of a zeolite BETA sample of 90% crystallinity obtained from batch
*
5'
58

The crystal size of zeolite BETA samples of the same crystallinity synthesized from
gels with different alkalinities is given in Table 6. At higher alkalinity larger
crystals are obtained. This is consistent with slower nucleation. The addition of
NaCl also increases the crystal size from 0.2 to 1 urn (B; compared to BI). The
batch B3 shows a clear effect of crystal growth by redissolution of the smallest
crystals. Upon extension of the crystallization period from 14 to 28 days, the
crystallinity decreases strongly (Fig.17), but the crystal size increases from 1 to
4 urn (Table 6). In this highly alkaline solution only the largest crystals are
stable. Because their rate of formation is slow an overall drop of the degree of
crystallinity is observed.
For batch B2 the crystal size was measured at 80 and 93% crystallinity. This
change in crystallinity is a very slow process since it involves 26 days of
synthesis (Table 6 and Fig.17).

TABLE 6. Effect of OH-/Si02 and Naf/SiC12 molar ratio on crystal size of zeolite BETA

Sample 0H/Si02 Na+/Si02 time(clays) crystallinity(X) crystal size(um)

Bl 0.52 0.10 28 82 0.1-0.2

52 0.54 0.12 4 80 0.2-0.3

B2 0.54 0.12 30 93 0.3


*
6.51 0.20 28 90 0.5-1.5
El

B3 0.62 0.20 14 59 1

B3 0.62 0.20 28 29 4

A small increase of crystal size accompanies these changes (Table 6). Since the
yield of crystals increases also from 42 to 5U% (Fig.5), it seems that no new
nuclei have been formed after reaching 80% crystallinity. As a result, at this
optimum alkalinity, the additional improvement of the crystallinity is due to the
very slow growth of existing crystallites.

4.2 Influence of the Si02/A1203 molar ratio

The influence of the Si02/A1203 molar ratio on the rate of crystallization has been
studied by adding various amounts of sodium aluminate to the batch of optimum
59

alkalinity. As shown in Fig.20, a decrease of the Al content apparently leads to an


increase in the crystallization rate. The effect can be better appreciated in
Fig.21, where the degree of crystallinity after 48 hours of synthesis has been
replotted against the aluminium molar fraction in the batches. Fourthy eight hours
corresponds to the stage of fastest crystallization for all the samples (Fig.20).
The yield of crystals is obtained by multiplying the degree of crystallinity by the
yield of solids (Fig.4) after 48 hours of synthesis. It is clear that there exists
an optimum Si02/A1203 ratio (between 60 and 100) in batches C and D, for which the
highest amount of crystals is obtained.

0 5 10 15 20 25 30

time /days

FIGURE 20. Influence of the Si02/A1203 molar ratio (values in brackets) of the
synthesis mixture on the degree of crystallinity of the solids as a
function of synthesis time.

These results are in contrast to what is reported for ZSM-5. For this zeolite
crystal growth continuously accelerates and nucleation becomes faster as the gel
becomes more siliceous 120, 22, 34-361.
In the literature no satisfactory explanation has been offered for the
frequently observed dependency of the crystallization kinetics on the Si02/A1203
ratio of the synthesis mixture 1341. The few explanations advanced always refer to
a pH drop in the system upon addition of aluminium 121, 341. Although this
explanation could be a priori considered with Al sources such as alumina, aluminium
hydroxide or cationic aluminium salts, it is not satisfactory with sodium
aluminate. The aluminate anion requires only two molecules of water to form Al(
60

which is the predominant species above a pH of 10 1431

AlO; + 2 H20 L, Al(OH&

0.05 0.10

Al /Al +Si gel

FIGURE 21. Degree of crystallinity of the solids and yield of crystals recovered
after 48 h synthesis as a function of the Al fraction in the synthesis
mixture (data from Fig.20).

In the system used here, consumption of OH- occurs only during the neutralisation
of silicic acid, resulting from the hydrolysis of tetraethyl orthosilicate. Since
for samples A, 62, C and D the OH-/Si02 ratio is the same (Table l), a pH effect
upon addition of sodium aluminate is absent and cannot account for the observed
differences. Considering the synthesis mechanism of Fig.16 and knowing that either
the rate of dissolution of the amorphous gel particles or the formation of adequate
precursors is rate determining, the existence of an optimum Al content for the
moment can only be explained by a change of rate-limiting step in the crystal
growth process. At lower aluminium fractions, there is evidence that the
dissolution of aluminium species from the amorphous solid is rate determining. For
higher Al fractions in the gel, this rate will be enhanced since all other
parameters remain unchanged. If instead the formation of adequate precursor species
is slowed down, the rate of nucleation and crystal growth could decrease. A high
resolution NMR study is in its way to verify this speculative explanation.
The crystal size of the zeolite BETA samples of roughly the same degree of
crystallinity but obtained from synthesis mixtures with different amounts of silica
are given in Table 7. Larger crystals are obtained from Al-rich systems, as is
expected when their slower nucleation is considered (Fig. 20).
61

4.3 Influence of TEA/Si02 molar ratio

The effect of the TEA/Si02 molar ratio on the crystallization rate has been
studied at Si02/A1203 = 30. Since TEA is added as TEA-OH, the alkalinity had to be
kept constant by correcting the NaOH levels.
By reducing the amount of TEA with respect to SiO2 at constant alkalinity both the
rate of nucleation and crystallization decrease (Fig.22). This indicates that TEA
plays an important role in the solution chemistry leading to the formation of
zeolite BETA.

TABLE 7. Effect of Si02/A1203 ratio on the crystal size of zeolite BETA

Sample Si02/A1203 time (days) degree of crystal size


in gel crystallinity (%) (um)

A 20 14 77 0.5-l
30 30 93 0.3
B2
C 60 14 95 0.2
D 100 4 99 0.2

0 5 10 15 20 25 30
TIME /DAYS

FIGURE 22. Influence of the TEA/Si02 molar ratio of the synthesis mixture on the
degree of crystallinity of the solids as a function of the synthesis
time (notations of Table 1).
62

From Table 8, which contains data on the crystal size of the samples, an increase
in crystal size with decreasing TEA content is obvious. This confirms the

nucleation enhancing effect of an increased TEA concentration. The data do not

allow to distinguish between a true template effect or a pore filling mechanism to

explain this behaviour. Anyway, the data of this and earlier sections clearly
indicate that the presence of TEA is essential for zeolite BETA synthesis.

TABLE 8. Influence of the TEAOH/Si02 ratio on crystal size of zeolite BETA

SAMPLE TEA/Si02 Time(days) degree of Crystal size


crystallinity (%) (vm)

0.50 30 93 0.3
B2
*
0.45 28 100 0.3 - 0.5
B2

ACKNOWLEDGMENTS
J. Perez-Pariente is grateful to K.U. Leuven (Belgium) and C.S.I.C.(Spain) for a

research grant. J.A. Martens and P.A. Jacobs acknowledge N.F.W.O. (Belgium) for a
grant and a research position, respectively. The authors also acknowledge the
sponsoring of this work by the Belgian Government (Dienst Wetenschapsbeleid) in the
frame of a concerted action on shape selective catalysis.

REFERENCES
1. K.L. Wadlinger, G.T. Kerr, E.J. Rosinski, U.S. Pat. 3,308,069 (1967) and

reissued U.S. Pat. Re. 28,341 (1975), assigned to Mobil Oil Corp.
2. J-A. Martens, M. Tielen, P.A. Jacobs and J. Weitkamp, Zeolites, 4 (1984) 98.
3. J.A. Martens, J. Perez-Pariente and P.A. Jacobs in "Chemical Reactions in
Organic and Inorganic Constrained Systems", R. Setton, ed., NATO AS1 Ser. C 165,
(1985) 115.
4. J.A. Martens, J. Perez-Pariente and P.A. Jacobs, Proceed. Int. Symp. Zeolite
Catalysis, Siofok (Hungary), 1985, Act. Phys. Chem. Szegedensis (1985) 487.
5. J.A. Martens and P.A. Jacobs, Zeolites 5 (1986),334.
6. L.B. Young, Eur. Pat. Appl. 30,084 (1981), assigned to Mobil Oil Corp..
7. M.A. Tobias, lJ.S.Pat. 3,728,408 (1973), assigned to Mobil Oil Corp..
8. R.B. La Pierre and S.S.F. Wong, Eur. Pat. Appl. 95, 304 (1983), assigned to

Mobil Oil Corp..


63

9. R-B- La Pierre and R.D. Partridge, Eur. Pat. Appl. 94,827 (Ig83), assigned to
Mobil Oil Corp..

10. R-B- La Pierre, R.D. Partridge, N.Y. Chen and S.S.F. Wong, Eur. Pat_ Appl.
95,303 (1983), assigned to Mobil Oil Corp..

II- R-B. La Pierre, RD. Partridge, N.Y. Chen and S.S.F. Wong, U.S. Pat. 4,5OI,g26
(I983), assigned to Mobil Oil Corp..

12. A- Erdem and l--B. Sand, Proceed. 5th Int. Conf. Zeolites, L.V.C. Rees, ed.,
Heyden, London (1980) 64.

13. F- Roozeboom, H.E. Robson and S.S. Chan, Zeolites, 3 (1983) 321.
14. C-L. Angel1 and W.H. Flank, A.C.S. Symp. Ser., 40 (1977) Ig4.
15. F. Polack and A. Cichocki, Adv. Chem. Ser., 121 (1973) 209.

16. S. Ueda, N. Kageyama and M. Koizumi, Proceed. 6th Int. Conf. Zeolites, D. Olson
and A. Bisio, eds., Butter-worth, Guildford (1984) 905.
17. S.P. Zhdanov, in "Molecular Sieves", The Society of Chemical Industry, London
(1968) 62.

18. S.P. Zhdanov, Adv. Chem. Ser., 121 (1973) 19.


19. F.E. Schwochow and G.W. Heinze, Adv. Chem. Ser., 101 (1971) 102.

20. E. Moretti, G. Leofanti, M. Padovan, M. Solari, G. De Alberti and F. Gatti,


Proceed. 8th Int. Congr. Catalysis, vol.4 (1984) 173.

21. P-A. Jacobs, E.G. Derouane and J. Weitkamp, J.C.S. Chem. Comm. (1981) 591
22. V.N. Romannikov, V.M. Mastikhim, S. Hocevar and B. Drzaj, Zeolites 3 (1983) 311.
23. Z. Gabelica, N. Blom and E.G. Derouane, Appl. Catal., 5 (1983) 227.

24. E.G. Derouane, S. Detremmerie, Z. Gabelica and N. Blom, Appl. Catal., 1 (1981)

201.

25. E.M. Flanigen, J.M. Bennett, R.W. Grose, J.P. Cohen, R.L. Patton, R.M. Kirchner

and J.V. Smith, Nature 271 (1978) 512.


26. D.M. Bibby, N.B. Milestone and L.P. Aldridge, Nature, 2BD (1979) 664.

27. R.M. Barrer, J.W. Baynham, F.W. Bultitude and W.M. Meier, J. Chem. Soc.,(I959)

195.
28. L.M. Parker, D.M. Bibby and J.E. Patterson, Zeolites, 4 (1984) 168.
29. V. Penchev, Ch. Minchev, V. Kanazirev, 0. Pencheva, N. Borisova, L- Kosovay H.

Lechert and H. Kacirek, Zeolites, 3 (1983) 249.


30. A. Araya, J--J.Barber,B.M. Lowe, D.M. Sinclair and A. Varma, Zeolitesp 4 (1g84)

263.
31. J-L. Cast-, T.V. Whittam and B.M. Lowe, Proceed. 6th Int. Conf. Ze"litesy D*

Olson and A. Bisio, Eds., Butterworth, Guildford (1984) 894.


32. B.E. Conway, R.E. Vera11 and J.E. Denoyers, Trans. Faraday Sac-, 62 (1966) 2738*

33. Z. Gabelica and E.G. Derouane, A.C.S. SYmP. Ser., 248 (1985) 21g-
34. R. Mostowicz and L.B. Sand, i!eOlites, 2 (1982) 143.
35. M. Ghamami and L.B. Sand, Zeolites, 3 (1983) 155.
36. K-J. Chao, T.C.Tasi, and M.S. Chen, J. Chem. Sot. Faraday Trans. IY 77 (lg81)

547.
64

37. S.B. Kulkarni, V.P. Shiralkar, A.N. Kotasthane, R.B. Borade, and P.Ratnasamy,
Zeolites, 2 (1982) 313.
38. G.T. Kerr, J.Phys.Chem., 70 (1966) 1047.
39. J. Ciric, Colloid. Int. Sci., 28 (1968) 315.
40. W. Meise and F.E. Schochow, Adv. Chem. Ser., 121 (1973) 169.
41. R. Aiello, C. Colella, and R. Sersale, Adv. Chem. Ser., 101 (1971) 51.
42. D. Domine, and J. Quobex, "Molecular Sieves", The society of Chemical Industry,
London (1968) 78.
43. R.M. Barrer, Zeolites, 1 (1961) 130.

You might also like