Direct observation of growth defects in zeolite beta

Article abstract of DOI:10.1021/ja043948s

High-resolution transmission electron microscopy reveals linear, "double pore" defects in the important zeolite beta. Structural interpretation of these defects gives evidence for the mechanism by which the zeolite crystallizes. Copyright

Full text of DOI:10.1021/ja043948s

Published on Web 12/22/2004
Direct Observation of Growth Defects in Zeolite Beta
Paul A. Wright,*,† Wuzong Zhou,*,† Joaquin Pe´rez-Pariente,‡ and Mar Arranz‡
School of Chemistry, UniVersity of St. Andrews, Purdie Building, North Haugh, St. Andrews, Fife KY16 9ST, U.K.,
and Instituto de Catalisis y Petroleoquimica, CSIC, 28049 Cantoblanco, Madrid, Spain
Received October 5, 2004; E-mail: paw2@st-andrews.ac.uk (P.A.W.); wzhou@st-andrews.ac.uk (W.Z.)
Zeolite beta is one of the most useful high silica zeolites, its
interconnected large pore network and strong acidity giving it
special catalytic properties. The elucidation of its structure solved
a longstanding problem in zeolite crystallography.1-3 Beta is highly
disordered, made up of a random intergrowth of two end-member
structures, polytypes A and B of Figure 1, defined by the mode of
stacking of complex silicate chain units. Each chain unit is stacked
at right angles to chains in the layer below and parallel to other
Figure 1. Beta polytypes A (tetragonal, P4122, left) and B (monoclinic
C2/c, right). Selected units in consecutively stacked layers are shown in
blue.
chain units in the same layer. Large pore channels, delimited by
12-MR pores, run parallel to and between adjacent chain units. If
two different stacking directions were to occur within the same
layer, the frameworks of the two stacking variants would be unable
A and B. Figure 2a shows regions in which the stacking vector
to connect at the boundary. This was taken to explain the observed
(equal to 1/3 of the in-layer repeat) is in the same direction for seven,
crystal morphology and also the anomalously large number of defect
eight and nine successive layers, at least in projection. There are
silanols reported for zeolite beta samples.2 No direct observation
also smaller regions involving five layers where the projection of
of intracrystalline defects of this sort has been reported, however.
the stacking vector alternates regularly (as it does in polytype A).
Electron microscopy is the most powerful technique for elucidating
Closer inspection also reveals defects that result from adjacent parts
defect structures;4 here, we show HRTEM evidence for such defects,
of the same layer being related by opposite stacking vectors to a
shedding light on the crystallization mechanism of zeolite beta and
common layer below. This is most easily seen from the different
its physical properties.
stacking of the large pores (the large white areas in the image).
For the case where the displacements are in opposite directions,
A sample of pure silica beta was prepared via the fluoride route,
according to published procedures,5 using dimethyldibenzylammo-
the layers are observed to come back into registry after three added
nium as template. XRD confirmed the bulk material as beta; SEM
layers (Figure 2c). The structure of such defects can be modeled
showed truncated tetragonal bipyramidal crystals 30 µm long with
in more detail by starting from a layer typical of the beta structure
and then adding building units with the two different stacking
some amorphous silica. HRTEM was performed using a JEOL
JEM-2011 electron microscope operating at 200 kV, with a point
orientations next to each other. The stacking direction perpendicular
resolution of 0.19 nm. The sample is electron beam sensitive, so
to the plane of the paper for this defect is assumed to be in the
the microscopic conditions were pretuned and orientation of the
same direction for the two domains over this region. This is likely
crystal was adjusted at low magnification until a principal zone
since short range order is common. The next layer cannot be
axis was found. Images were recorded at 600 000× using a Gatan
completed, but continued growth of the two regions with the same
two vectors does enable the subsequent layer to overgrow (Figures
2d and 3). Growth onto this layer can then take place with either
794 CCD camera at very low beam brightness.
HRTEM (Figure 2) reveals the disordered stacking sequence of
layers (and 12-membered ring pores) characteristic of zeolite beta.
stacking direction.
Beta is made up of layers of chain units stacked perpendicular to
The interrupted structure results in Q3 silicons (Si(-OSi-)3OH)
lining the defect as shown in the model, which were made using a
kit with atomic centers of appropriate geometry (Figure 3). Silanol
hydroxyls are known to be a marked feature of beta materials.2
The model shows remarkable agreement with the image. The
predicted arrangement of two pores, double and one-and-a-half
those of the layer below, a regular distance (a) apart, so that each
one is translated by (a/3 relative to the unit below. There is
therefore a family of possible end-members with different stacking
sequences,6 including polytypes A (chain stacking vectors +a/3;
-a/3; +a/3, etc.) and B (stacking +a/3; +a/3, ... along the related
direction). The projections (along a common direction relative to
times as large as the usual large 12-MR pore, is mirrored.
the layers of the two polytypes, i.e., [100] for polytype A) are shown
Furthermore, careful inspection of the secondary detail is also in
in Figure 1. Projections do not distinguish between polytypes which
agreement. Fourier averaging of a “perfect” region of the structure
have different stacking vectors for the chain units running parallel
gives the image in Figure 2b. At this resolution, it is possible to
to the plane of the paper. It is similarly not possible from TEM
resolve, in addition to the large pores, the areas of low projected
images to determine the direction of stacking perpendicular to the
electron density that result from six-, five-, and even four-membered
page. However, the image (Figure 2) does reveal nanodomains
rings. Building units made up from four 5-MRs and one 6-MR in
ordered in two dimensions. The microstructure observed here is
projection (outlined in Figure 1) are observed to stack in the same
similar to that of typical samples of beta1,2 or its natural counterpart,
orientation. Although the image of the defects (Figure 2c) is noisier
disordered tschernichite,7 with a structure related to end-members
(their irregular distribution prevents Fourier averaging), it is possible
to make out the orientation of building units of this type. Going
through each defect along a layer, we observe that the orientation
† University of St. Andrews.
‡ Instituto de Catalisis y Petroleoquimica.
9
494
J. AM. CHEM. SOC. 2005, 127, 494-495
10.1021/ja043948s CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. (a) HRTEM of an edge of a beta crystallite, annotated to show the stacking directions in different parts. Nanodomains related to polytypes A and
B are indicated. Defects are visible in the center of the image, where domains with different stacking directions meet. (b) Fourier-averaged image of a
domain of type B, with structural units outlined. (c) Image of two double pore defects, showing different stacking directions and outlining secondary structural
details on either side of the defects. (d) Suggested crystal growth sequence, indicating how the defects form and how the building units adopt different
orientations at the defect (similar subunits are colored in similar colors; only the pale unit at the core of the defect cannot unambiguously be identified in
the image).
tionalized siloxanes can be incorporated within the pores of beta
during synthesis, without loss of order, giving shape-selective,
organic-functionalized microporous catalysts. It is tempting to
speculate that organic groups could be incorporated at defects
similar to those described here. Finally, the defects suggest a way
to include well-defined mesoporosity within a zeolite if nucleation
rates can be controlled by changing synthesis conditions.
Acknowledgment. CICYT (MAT 2003-07769.C02-02) and the
Spanish Ministries of Science (M.A.) are thanked for funding. Dr.
T. Blasco (ITQ, Valencia) is thanked for the NMR spectra.
Supporting Information Available: XRD and 29Si MAS NMR.
This material is available free of charge via the Internet at http://
pubs.acs.org.
Figure 3. Model of the observed defects, obtained by stacking in the two
ways onto a single layer and continuing to the third similar layer. The cages
indicated in light pink in Figure 2d may or may not be complete. Here
they are not, whereas they are in the model shown in the graphical abstract.
References
of these units reverses. The exact nature of the core of the defect
(1) Treacy, M. M. J.; Newsam, J. M. Nature 1988, 332, 249.
(2) Newsam, J. M.; Treacy, M. M. J.; Koetsier, W. T.; de Gruyter, C. B.
remains unclear, even at such high resolution; models are possible
with or without [54‚43] cages building out into the pores. Computa-
Proc. R. Soc. A 1988, 420, 375-405.
(3) Higgins, J. B.; LaPierre, R. B.; Schlenker, J. L.; Rohrman, A. C.; Wood,
tion studies may resolve the issue.8 The linear defects (perpendicular
J. D.; Kerr, G. T. Zeolites 1988, 8, 446.
to the plane of projection) run through the structure. The two larger
pores are partly partitioned by hydroxyls, but the space within the
pore system is enlarged, particularly at the intersections of the
defects with channels running perpendicular.
Observation of these defects casts light on two intriguing aspects
of the synthesis and growth of zeolite beta. First, they are consistent
with a layer-by-layer growth mechanism for beta, where growth
may nucleate at different parts of the same layer with different
stacking vectors. Such a growth mechanism is shown schematically
in Figure 2d, indicating how the defects could heal. A similar
mechanism has been suggested for the double defects observed in
ETS-10 by Anderson et al.9 Jones et al.10 have shown that func-
(4) Thomas, J. M.; Terasaki, O.; Gai, P. L.; Zhou, W.; Gonzalez-Calbet, J.
Acc. Chem. Res. 2001, 34, 583.
(5) Arranz, M.; Garcia, R.; Pe´rez-Pariente, J. In Proceeding of the 14th
International Zeolite Conference, South Africa, 2004; p 256.
(6) Gies, H.; van Koningsveld, H. Catalogue of Disorder in Zeolite
Frameworks (http://www.iza-structure.org/database/).
(7) Szostak, R.; Pan, M.; Lillerud, K. P. J. Phys. Chem. 1995, 99, 2104.
(8) Ohsuna, T.; Slater, B.; Gao, F.; Yu, J.; Sakamoto, Y.; Zhu, G.; Terasaki,
O.; Vaughan, D. E. W.; Qiu, S.; Catlow, C. R. A. Chem.sEur. J. 2004,
5031.
(9) Anderson, M. W.; Agger, J. R.; Hanif, N.; Terasaki, O. Microporous
Mater. 2001, 48, 1.
(10) (a) Jones, C. W.; Tsuji, K.; Davis, M. E. Nature 1998, 393, 52. (b) Jones,
C. W.; Tsapatsis, M.; Okubo, T.; Davis, M. E. Microporous Mater. 2001,
42, 21.
JA043948S
9
J. AM. CHEM. SOC. VOL. 127, NO. 2, 2005 495