Method for the in situ preparation of a single layer of zeolite Beta crystals on a molybdenum substrate for microreactor applications

Article abstract of DOI:10.1016/j.jcat.2007.02.007

A method for the hydrothermal synthesis of a single layer of zeolite Beta crystals on a molybdenum substrate for microreactor applications has been developed. Before the hydrothermal synthesis, the surface of the substrate was modified by an etching procedure that increases the roughness at the nanoscale level without completely eliminating the surface lay structure. Then, thin films of Al₂O₃ (170 nm) and TiO₂ (50 nm) were successively deposited by atomic layer deposition (ALD) on the substrate. The internal Al₂O₃ film protects the Mo substrate from oxidation up to 550 °C in an oxidative environment. The high wettability of the external TiO₂ film after UV irradiation increases zeolite nucleation on its surface. The role of the metal precursor (TiCl₄ vs TiI₄), deposition temperature (300 vs 500 °C), and film thickness (50 vs 100 nm) was investigated to obtain titania films with the slowest decay in the superhydrophilic behavior after UV irradiation. Zeolite Beta coatings with a Si/Al ratio of 23 were grown at 140 °C for 48 h. After ion exchange with a 10^-4 M cobalt acetate solution, the activity of the coatings was determined in the ammoxidation of ethylene to acetonitrile in a microstructured reactor. A maximum reaction rate of 220 μmol C₂H₃N g^-1 s^-1 was obtained at 500 °C, with 42% carbon selectivity to acetonitrile.

Full text of DOI:10.1016/j.jcat.2007.02.007

Journal of Catalysis 247 (2007) 328–338
www.elsevier.com/locate/jcat
Method for the in situ preparation of a single layer of zeolite Beta crystals
on a molybdenum substrate for microreactor applications
M.J.M. Mies a, E.V. Rebrov a, J.C. Jansen b, M.H.J.M. de Croon a, J.C. Schouten a,∗
a
Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands
b
Catalysis Engineering, DelftChemTech, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands
Received 2 October 2006; revised 30 January 2007; accepted 4 February 2007
Abstract
A method for the hydrothermal synthesis of a single layer of zeolite Beta crystals on a molybdenum substrate for microreactor applications has
been developed. Before the hydrothermal synthesis, the surface of the substrate was modified by an etching procedure that increases the roughness
at the nanoscale level without completely eliminating the surface lay structure. Then, thin films of Al O (170 nm) and TiO (50 nm) were
2
3
2
successively deposited by atomic layer deposition (ALD) on the substrate. The internal Al O film protects the Mo substrate from oxidation up to
2
3
◦
550 C in an oxidative environment. The high wettability of the external TiO film after UV irradiation increases zeolite nucleation on its surface.
The role of the metal precursor (TiCl vs TiI ), deposition temperature (300 vs 500 C), and film thickness (50 vs 100 nm) was investigated to
2
◦
4
4
obtain titania films with the slowest decay in the superhydrophilic behavior after UV irradiation. Zeolite Beta coatings with a Si/Al ratio of 23
◦
−4
were grown at 140 C for 48 h. After ion exchange with a 10 M cobalt acetate solution, the activity of the coatings was determined in the
−1 −1
was obtained at
ammoxidation of ethylene to acetonitrile in a microstructured reactor. A maximum reaction rate of 220 µmol C H N g
500 C, with 42% carbon selectivity to acetonitrile.
s
2
3
◦
© 2007 Elsevier Inc. All rights reserved.
Keywords: Zeolite Beta; In situ synthesis; Thin zeolitic coatings; Molybdenum; Atomic layer deposition; Surface roughness; Surface hydrophilicity; Protective
coatings; Microreactor; Ethylene ammoxidation; Acetonitrile
1. Introduction
tion units [13], catalytic packings [14,15], monoliths [16], and
deNOx reactors [17].
Three-dimensional large-pore zeolite Beta (BEA) is cur-
rently applied in various catalytic gas- and liquid-phase proces-
ses; important examples include the fluid catalytic cracking
and dewaxing of petroleum oils [1], alkylation and acyla-
tion of aromatics [2,3], isomerization of alkanes [4], Fischer–
Tropsch synthesis of iso-paraffins [5], deNOx [6,7], selective
hydrogenations [8,9], and ammoxidation of light paraffins and
olefins [10]. Besides these large-scale processes, zeolite Beta
also is applied in the synthesis of fine chemicals [3]. Extrudates
of zeolite Beta, consisting of zeolite crystals and a binder mate-
rial, are often applied, resulting in low catalyst efficiency [11].
Application of zeolitic coatings improves catalyst performance,
as demonstrated in the case of membrane reactors [12], distilla-
Fast catalytic reactions involving large heat effects can be
further intensified when carried out over zeolitic coatings up
to 5 µm thick grown directly in the channels of a microstruc-
tured reactor, fabricated in a metal substrate [18,19]. The highly
exothermic ammoxidation of ethylene to acetonitrile over zeo-
lite Beta coatings exemplifies such reactions. The large geo-
metrical surface area of the channel walls provides a large
interface area between the reactants and the coating. The ab-
sence of both binder and macropores improves mass transfer
in the coating [11,18]. Heat transfer is enhanced by the chemi-
cal bonds between the coating and the substrate, as well as the
higher apparent density of the coating [20]. Therefore, methods
need to be developed and optimized for the deposition of thin
zeolitic coatings on the metal substrate. The hydrothermal syn-
thesis method is an elegant way to grow a single layer of zeolite
crystals directly in microchannels [18,19,21].
*
Corresponding author.
E-mail address: j.c.schouten@tue.nl (J.C. Schouten).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.02.007

M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
329
Molybdenum is often chosen as a substrate material for
microreactor applications due to its high thermal conductiv-
ity and mechanical stability [22–24]. Above 350 ◦C in the
presence of oxygen, molybdenum oxidizes to orthorhombic
MoO3 [25]. Furthermore, molybdenum dissolves during the in
situ growth of zeolitic coatings in the highly alkaline synthesis
mixture [23]. Therefore, an approach to developing Mo-based
(micro-)structured devices is to use the intrinsic properties of
molybdenum through the application of thin protective films on
the metal surface [25–28]. Both chemical [29,30] and structural
[31,32] properties of the substrate surface, as well as its wetta-
bility, are of importance to in situ zeolite growth. The surface
roughness affects zeolite growth as well as bonding of zeolite
crystals with the substrate. A roll casting process used for metal
foil production usually produces a surface with lay (i.e., the pre-
dominant direction of the surface texture). Striations or peaks
and valleys on a metal foil are usually observed in the direction
in which the tool was drawn across the surface. The presence of
lay is an important factor in the hydrophilic properties of a sur-
face.
Substrate wettability can be improved by surface treatments
[33,34] or by deposition of a hydrophilic film [23]; for example,
a titania film can be made superhydrophilic (>15 OH nm−2) by
UV irradiation [35,36]. Titania thin films, deposited by atomic
layer deposition (ALD), have been the subject of extensive re-
search [37–39] due to different applications such as fuel cells,
chemical sensors, and photocatalysis. Films that were grown
below 140 ◦C had an amorphous structure, whereas a poly-
crystalline structure appeared at deposition temperatures above
165 ◦C regardless of the crystallinity of the substrate [37]. Films
grown above 210 ◦C revealed a polycrystalline anatase structure
for a film thickness of 15–55 nm. The influence of the substrate
on the TiO2 film growth has been investigated on substrates in-
cluding Si [40], KBr [41], soda-lime glass [42], and MgO [43].
In all cases, anatase was formed at 300 ◦C with no preferred ori-
entation. The deposition of ALD titania films has been focused
primarily on using the metal sources titanium chloride, titanium
isopropoxide, and titanium ethoxide. Water is the primary non-
metal precursor used for titania growth. More recently, titanium
iodide has been investigated as a possible precursor [44–46].
Compared with the more common TiCl4-based ALD process,
the TiI4-based process yields a growth rate about four times
higher on α-Al2O3 substrates at 600 ◦C. Furthermore, the TiI4
process was found to produce films with a higher degree of
crystallinity at lower temperatures. Therefore, the advantage of
TiI4 compared with TiCl4 seems to be its relatively easy ligand
release at temperatures above 300 ◦C. Films grown at 400 and
500 ◦C had a mixed anatase and rutile crystal structure, with ru-
tile the dominant phase. Precursors and deposition techniques
also contribute to the phase formation.
ammoxidation to acetonitrile in a microstructured reactor was
evaluated.
2. Experimental
2.1. Surface modification of the molybdenum substrate
Experiments were carried out on 10 × 10 mm2 and 40 ×
10 mm2, 100-µm-thick molybdenum (Aldrich, 99.9+ wt%)
substrates. The substrates were boiled in xylene for 1 h to
remove organic contaminations, dried at 140 ◦C for 1 h, and
treated by an etching mixture containing 4.3 vol% H2O2
(Fluka) and 3.6 vol% NH4OH (Riedel de Haën) in distilled wa-
ter. Then aluminum and titanium oxide thin films were succes-
sively deposited by ALD on the Mo substrates. The deposition
of the films was performed in a flow-type reactor of 0.196 L
at a pressure of about 5 mbar. Trimethylaluminium (TMA)
was used as the metal precursor for the alumina film, and ti-
tanium chloride (TiCl4) or titanium iodide (TiI4) was used for
the titania film. In all cases, H2O was used as the oxidant. The
H2O, TMA, TiCl4, and TiI4 precursors were stored in reser-
voirs outside the reactor at 22, 20, 22, and 110 ◦C, respectively.
In the TMA–H2O and TiCl4–H2O processes, precursor pulses
of 600 ms were separated by nitrogen pulses of 1200 ms. In the
TiI4–H2O process, the following pulse sequence was applied:
TiI4 (2500 ms)–N2 (2500 ms)–H2O (600 ms)–N2 (1200 ms).
A total of 2431 cycles was required for the deposition of a 170-
nm alumina film. The deposition of 50- and 100-nm titania
films required a number of 1111 and 2222 cycles, respectively.
Before zeolite synthesis, the substrates were UV irradiated for
3 h at room temperature in a metal box (UV lamp: Hanovia
679A-36; 450 W; λ range, 220–1400 nm). Hereinafter, the
TiO2/Al2O3/Mo substrate is referred to as the TAMo substrate.
2.2. Zeolite Beta precursor gel
The silica and alumina solutions were prepared separately.
The silica solution was prepared by dissolving appropriate
amounts of NaCl (Merck), KCl (Merck), and tetraethylammo-
nium hydroxide (TEAOH, 40 wt%, Fluka) in demineralized
water, after which colloidal silica (Ludox HS-40, 40 wt% in
water) was added. The silica solution was stirred for 0.5 h until
the mixture was homogenized. The alumina solution was pre-
pared by dissolving NaOH (Merck) and NaAlO2 (0.53 wt%
Al2O3, 0.44 wt% Na2O, Riedel de Haën) in demineralized wa-
ter under continuous stirring. Subsequently, the silica solution
was slowly added to the alumina solution, after which the mix-
ture was shaken vigorously to prevent formation of a gel. The
synthesis mixture was stirred at high speed (600 rpm) for 1 h
at room temperature. Gels of the following oxide molar com-
position were prepared for the synthesis: 1.6 Na2O:0.93 K2O:
11.4 (TEA)2O:1 Al2O3:46 SiO2:710 H2O:1.8 HCl.
The present study focused on the development of a method
for the hydrothermal growth of zeolite Beta coatings with a sin-
gle layer of crystals on a Mo substrate. First, the surface of
the molybdenum substrate was modified before hydrothermal
synthesis. Then, synthesis conditions were optimized to yield
a perfectly ordered single layer of zeolite Beta crystals. Fi-
nally, the activity of the Co-zeolite Beta coatings in the ethylene
2.3. Hydrothermal treatment
Directly after the UV irradiation step, the TAMo substrate
was positioned in the synthesis mixture at 20 mm below the

330
M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
taking the weight loss per area (g m−2) after the etching proce-
dure. For example, a substrate that lost 4 g m−2 after etching is
referred to as S4, whereas the initial substrate is designated S0.
Film hydrophilicity was evaluated by contact angle measure-
ments of water on the titania films using an automatic con-
tact angle meter (DataPhysics OCA 30) at a relative humidity
of 40%. A 2.0-µL droplet of demineralized water was dropped
on the surface of the substrate. The contact angle of the water
droplet with the substrate was determined with SCA 20 soft-
ware before the UV irradiation and 10 and 200 min afterward.
The weight difference of the TAMo substrate before and the
TAMo substrate with the zeolitic coating after the hydrother-
mal synthesis was determined with an analytical mass balance
(Mettler Toledo XS105) and used to calculate the zeolite cover-
age. The Si conversion of the synthesis mixture is defined as
Fig. 1. (a) A 2D surface profile of a substrate showing the intrinsic waviness.
(b) The average roughness value, R , is calculated from the 2D profile after
a
correction of the waviness. The R value is calculated from the integral of
the absolute value of the roughness profile (shaded area) divided by the length
a
mol(Sipowder + Sicoating
)
Siconversion
=
× 100%,
(2)
L (Eq. (1)). (c) The R value does not provide information about the surface
molSiprecursor
a
roughness at the nanoscale level.
where the molar amounts of Sipowder and Sicoating are calcu-
lated from both the as-synthesized zeolite powder and coatings
after correction for the amount of water, template, and cations
present in the zeolite.
gas–liquid interface with 10 × 10 mm2 surfaces parallel to the
gravity vector. The synthesis was carried out at 140 or 150 ◦C
under autogenic pressure for up to 48 h. A 50-mL PEEK in-
sert filled with 35 mL of solution was positioned in a stainless
steel outer shell of the autoclave and closed. In some experi-
ments, the outer shell of the autoclave was preheated to 180 ◦C
to provide rapid heating of the synthesis mixture with an ini-
tial heating rate of 10 ◦C min−1 [47]. The zero of synthesis
time was taken as the time when the autoclave was placed in
a convection oven maintained at the synthesis temperature. The
syntheses were performed at static conditions. After the synthe-
sis, the autoclave was quenched in water to room temperature.
The zeolite particles from the bottom of the autoclave were
filtered through a 0.22-µm filter (Millipore), washed with dem-
ineralized water, and dried for 24 h at room temperature.
The synthesized coatings were examined by X-ray diffrac-
tion (XRD) for phase identification and crystal orientation.
XRD data were collected on a Rigaku Geigerflex diffractometer
using CuKα radiation (1.5405 Å). XRD patterns were recorded
in the range of 5–50◦ 2θ using step scanning at 0.02◦ 2θ per
step and a counting time of 4 s for each step. The coating sur-
face coverage and crystal morphology were evaluated by scan-
ning electron microscopy (SEM) using a JEOL JSM-840A mi-
croscope and atomic force microscopy using a Spema-B scan-
ning probe microscope with Nova RC1 software. The crystal
size distribution and the average crystal size were determined
from SEM images of various positions on the surface, where the
total number of analyzed crystals was at least 300. The elemen-
tal composition of the coatings was analyzed by X-ray photo-
electron spectroscopy (XPS) on a VG Escalab MKII spectrom-
eter, equipped with a dual Al/MgKα X-ray source and a hemi-
spherical analyzer with a five-channeltron detector. Spectra
were obtained using the aluminium anode (AlKα = 1486.6 eV)
operating at 250 W and a constant pass energy of 20 eV with
a background pressure of 2 × 10−9 mbar. Nitrogen adsorption
isotherms were obtained on an ASAP-2020 Micromeritics in-
strument with a standard procedure after vacuum pretreatment
at 300 ◦C for 12 h up to a residual pressure below 0.1 Pa. The
surface area of the zeolitic coating was determined from the ni-
trogen adsorption–desorption isotherms at −196◦C in the range
of relative pressures of 0.001–0.4. The pore size distribution
was obtained by the Horvath–Kawazoe method. The full pore
volume was calculated from the maximum adsorption value ob-
tained from the N2 adsorption isotherm.
2.4. Characterization of substrate and coatings
The substrate surface roughness (Ra) was determined by
laser scanning confocal microscopy (LSCM). The digitized
two-dimensional (2D) surface profile (Fig. 1a) was corrected
for the intrinsic waviness of the substrate (Fig. 1b). The aver-
age Ra values were determined by Eq. (1) from three 2D height
profiles taken at three different locations on the substrate paral-
lel to the direction of the lay of the surface,
L
ꢀ
1
Ra =
|r(x)|dx,
(1)
L
0
where r(x) is the difference between the absolute height and
the least squares mean line of the 2D surface profile at posi-
tion x and L is the total length of the analyzed profile (Fig. 1b).
Ra does not include information about the frequency that the
height profile crosses the least squares mean line. Furthermore,
Ra does not include information about the roughness of the
surface at the nanoscale level (Fig. 1c). Therefore, a second pa-
rameter was introduced to characterize the surface roughness by
2.5. Ethylene ammoxidation reaction
The as-synthesized coatings were subjected to a posttreat-
ment procedure consisting of calcination and ion-exchange
steps to obtain Co(II)-zeolite Beta coatings. The sequence and

M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
331
Table 1
Overview of the post-treatment steps used for coating activation
Step
no.
Process
Temperature
( C)
Heating rate
Medium
Other
conditions
Repetition
Dwell
(h)
◦
◦
−1
( C min
)
1
2
3
4
5
6
Water rinse
Ultrasonic treatment
Drying
Calcination
Calcination
Ion-exchange
r.t.
r.t.
110
295
500
80
–
–
–
1
1
–
Demi water
Demi water
Air
Air
Helium
–
–
–
–
–
–
3
–
1
12
2
2
2
60 Hz, 120 W
–
50 mL min
50 mL min
Reflux cooler,
−1
−1
1 M NH NO
4
3
stirring
7
8
9
Water rinse
Drying
Calcination
Ion-exchange
80
110
500
80
–
–
1
–
Demi water
Demi water
–
–
2
–
–
–
0.75
12
2
−1
Helium
100 mL min
Reflux cooler,
stirring
–
–
−4
10
10 M cobalt
acetate, 200 mL
Demi water
Demi water
Air
3
11
12
13
14
Water rinse
Drying
Calcination
Calcination
r.t.
110
295
500
–
–
1
1
2
–
–
–
–
12
2
−1
50 mL min
50 mL min
−1
Helium
2
conditions of the various treatments are specified in Table 1.
XPS analysis on the coatings after the posttreatment resulted in
a Si/Al ratio of 23 and a Co/Al ratio of 0.5.
Table 2
Etching conditions and surface roughness of the TAMo substrates
TAMo
Conditions
Surface parameters
R (nm) ꢀw (g m
a
substrate
◦
−2
The reaction was carried out in a molybdenum plate-type
microreactor. Four coated TAMo substrates with dimensions of
40 × 10 × 0.1 mm3 were inserted in a microstructured reactor
at a distance of 350 µm from one another. The reactor was op-
erated in the differential mode [22]. Blank experiments were
carried out on inert titanium plates to exclude the contribution
of the reactor material to reactant conversion. A total catalyst
weight of 0.0018 g was applied for the reaction. The Co-Beta
coatings were pretreated in a helium flow of 33 mL min−1 for
1 h at 530 ◦C. Subsequently, the reactor temperature was de-
creased to the desired value, and the reactant mixture was intro-
duced at a flow rate of 33 mL min−1 (STP) with a composition
of ethylene (2 vol%), ammonia (2 vol%), and oxygen (2 vol%),
balanced by helium. The reaction rate and product selectivity
were determined under steady-state conditions at 350–500 ◦C,
with a 1-h dwell at each temperature. The concentration of the
components in the effluent gas was analyzed with an online Var-
ian micro-gas chromatograph (CP-4900), equipped with a ther-
mal conductivity detector. Helium was used as a carrier gas.
The analysis time was 80 s for each injection. N2, O2, CH4, NO,
and CO were analyzed on a molsieve 5A column (0.25 mm ID,
10 m) operated at 175 ◦C and 350 kPa. C2H4, C2H6, CO2, NH3,
and H2O were analyzed on a poraPLOT-U column (0.25 mm
i.d., 10 m long) operated at 65 ◦C and 200 kPa. CH3CN and
CH3OH were analyzed on a poraPLOT-U column (0.25 mm
i.d., 10 m long) operated at 175 ◦C and 150 kPa. The carbon
and nitrogen mass balances were closed within 95%; however,
the carbon balance was closed only under the assumption that
methanol was formed in equimolar amounts with another C1-
containing component (product Y), which could not be ana-
lyzed separately with the micro-gas chromatograph.
T ( C)
t (s)
)
a
S
S
S
S
–
–
13
14
45
90
2
5
9
8
0
4
55
0
50
70
80
30
60
120
4
55
120
a
120
The subscript indicates the specific weight loss ꢀw of the Mo substrate after
the etching treatment with a mixture of 4.3 vol% H O (Fluka) and 3.6 vol%
2
2
NH OH (Riedel de Haën) in distilled water.
4
gen selectivity (S ) to a nitrogen-containing product, Pj , are
Nj
defined as
y
· n
Ci
Ci
ꢁ
SC
=
· 100%
(3)
(4)
i
y
· n
Ci
Ci
i
and
y
· m
Nj
Nj
ꢁ
SN
=
· 100%,
Nj
j
y
· m
Nj
j
where yC and yN are the molar fractions and nC and mN
i
j
i
j
are the number of carbon and nitrogen atoms in the product
molecule, respectively.
3. Results and discussion
3.1. Effect of surface roughness on coverage
The surface roughness of the substrate was increased by
a chemical etching procedure (Table 2). LSCM scans of the
TAMo substrates (S0, S4, S55, and S120) before the synthesis
are shown in Figs. 2a, 2c, 2e, and 2g. The corresponding SEM
images of the as-synthesized zeolite Beta coatings are shown in
Figs. 2b, 2d, 2f, and 2h. After the synthesis, uncovered areas
remained on the nontreated substrate S0, which has a constant
waviness profile obtained in a roll-casting process. A closed
layer of zeolite Beta crystals was obtained on S4, which has
The reaction rate is expressed either in terms of the acetoni-
trile formation rate per gram of zeolite per second or in terms of
the turnover frequency per single Co atom (TOF, s−1). Carbon
selectivity (SC ) to a carbon-containing product,Pi, and nitro-
i

332
M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
2
Fig. 2. LSCM scans (scan area is 320 × 320 µm ) of the TAMo substrates: (a) S , (c) S , (e) S , and (g) S
(Table 1) and the corresponding scanning electron
0
4
55
120
micrographs of the as synthesized zeolite Beta coatings (b, d, f, and h, respectively). The black arrows in the LSCM scans indicate the direction of the lay of the Mo
substrate parallel to which the 2D height profiles were digitized for the determination of the R value. Synthesis conditions: T = 150 C, t = 30 h, initial heating
◦
a
◦
−1
.
rate = 10 Cmin
a similar surface roughness but a different surface morphology.
Substrate S4 was modified by a mild etching treatment only at
the nanoscale level, which does not influence Ra (Fig. 1c). The
zeolite coverage was increased by 35%, which is attributed to
the increased number of surface irregularities. It was shown that
formation of silicalite-1 and zeolite Y was improved on a plas-
tically deformed or mechanically destroyed copper substrate
because of the higher number of surface imperfections, which
are supposed to be nucleation centers for zeolite crystals [31,
32]. It is expected that surface irregularities are far more homo-
geneously distributed over the surface after chemical etching
compared with a mechanical surface treatment. Compared to
S0, S4 has also a higher surface energy, which is favorable for
anchoring of small crystals. The average diameter of zeolite
crystals is 1.5 µm. It is characteristic of S0 and S4 morphology
that the difference in altitude between the highest and lowest
points amounts to 1.3 µm in the direction perpendicular to lay.
This is less than the mean diameter of a zeolite crystal of 1.5 µm
(see Fig. 3a).
Substrates S55 and S120 were subjected to more severe etch-
ing treatments (Table 2), which resulted in an increased surface
roughness (Figs. 2e and 2g). For S55 and S120, the maximum
difference between the top and bottom part of the substrate sur-
face of 2.0 and 2.9 µm, respectively, is larger than the average

M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
333
Fig. 3. Illustrations of the cross sections of the two-dimensional surface profiles of the zeolite Beta coatings on a Mo substrate. Panel (a) shows the surface profile of
the substrates S or S , on which a single layer of zeolite Beta crystals (average crystal size is 1.5 µm) are visible (compare (a) with Figs. 2a–2d). Panel (b) shows
0
4
the surface profile of substrate S
(Table 2) with a R value of 90 nm, consisting of large pockets in which zeolite Beta crystals can be deposited on top of each
a
120
other (compare (b) with Figs. 2g and 2h).
diameter of the zeolite crystals. Therefore, the second layer of
zeolite crystals can be deposited on top of the crystals located in
the substrate pockets, as shown schematically in Fig. 3b. Thus,
to obtain a single layer of zeolite crystals, the surface morphol-
ogy has to contain numerous imperfections, while the available
pockets should not exceed the size of the crystals.
3.2. Effect of surface hydrophilicity on coverage
The alumina and titania films were successively deposited
by ALD [27,48]. The inner alumina film consists of a non-
porous amorphous phase, which protects the substrate from
oxidation in an oxygen-containing mixture [22]. No reflections
from molybdenum oxides were observed in the XRD patterns
of a Mo substrate coated with a 170-nm thin alumina film after
10 h in an air–water mixture at 550 ◦C (Fig. 4b). The den-
sity of the alumina film of 3.5 g cm−3 is slightly lower than
for α-Al2O3 (3.97 g cm−3), whereas their surface properties
are reported to be similar [27]. Because the alumina film con-
tains less than 1 OH group per nm2 [33], an outer titania film
was deposited to increase the surface wettability of the sub-
strate [34].
Fig. 4. XRD patterns of substrates after various surface modifications and
treatments; (a) molybdenum (Mo) substrate, (b) Al O (170 nm)/Mo sub-
2
3
◦
strate after 10 h of oxidation in an air/water (2.3 vol%) mixture at 550 C,
(c) TiO (50 nm)/Al O (170 nm)/Mo substrate (TAMo substrate). The reflec-
2
2 3
tions Mo (2) and anatase TiO (") are of the crystalline phases for indicated
2
in the figure.
The deposition temperature, film thickness, and nature of
the titanium precursor (TiCl4 and TiI4) were varied to obtain
an optimal hydrophilicity of the TiO2 film after UV irradiation
(Table 3). The difference between the thermal expansion coeffi-
cient of film and substrate must be as small as possible to avoid
crack formation. For the same reason, thinner films are more
preferred [49]. The highest temperature in the zeolite posttreat-
ment is 500 ◦C. Therefore, the ALD films should preferably
be deposited at this temperature. Before the UV treatment, the
lowest water contact angle of 44◦ was observed for a film of
50 nm deposited at 500 ◦C from the TiCl4 precursor. This film
has a low density of surface OH groups because of thermal de-
hydroxylation of the surface, resulting in a rather hydrophobic
surface at this temperature. After the UV irradiation, superhy-
drophilic behavior (contact angle <5◦) was observed for all
samples. The films obtained from TiCl4 at 300 ◦C remained su-
perhydrophilic for at least 200 min, whereas those obtained at

334
M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
Table 3
ALD parameters and water contact angle on the S substrate
0
◦
Temperature Thickness
Pretreatment/titania precursor/contact angle ( )
a
◦
( C)
(nm)
No UV
UV (min)
TiCl
TiI
4
TiCl
10
TiI
4
4
4
200
10 200
b
300
500
50
100
50
59
68
44
77
3
75
4
<5
<5
<5
<5 <5
9
8
7
69 12 <5
<5
9
Fig. 5. Scanning electron micrographs of zeolite Beta coatings without (a)
74
78
9
5
<5 (<5) <5 (9) <5 28
<5 <5 30
and after (b) a UV irradiation of the S substrate. Synthesis conditions:
100
6
55
◦
◦
−1
.
T = 150 C, t = 16 h, initial heating rate = 1 Cmin
a
Time between UV treatment and contact angle analysis. The substrate was
positioned perpendicular to the UV source. Values in the parentheses are given
for the parallel orientation of the substrates relative to the UV source.
b
◦
Contact angle values below 5 cannot be analyzed accurately.
500 ◦C showed a slight increase of the contact angle. One may
conclude that the deposition temperature and the film thick-
ness do not influence the performance of the films obtained
in the TiCl4–H2O ALD process. However, the films obtained
from TiI4 became much more hydrophobic, especially those
deposited at 500 ◦C. This may be due to the different struc-
ture of the films. Both hydroxyl groups and chlorine ions can
be easily incorporated in the film structure when TiCl4 is used
as a metal precursor. The TiI4 process is faster than the TiCl4
process, due to relatively easy ligand release at temperatures
above 300 ◦C. Therefore, the TiI4 process produces films with
a higher degree of epitaxy and with a lower content of the bulk
OH groups; however, the uniform films obtained in the TiI4
process show a fast decay in hydrophilicity. This effect is es-
pecially pronounced at 500 ◦C. Therefore, TiCl4 was chosen
as an ALD precursor for deposition of 50-nm titania films at
500 ◦C. At these conditions, the anatase structure with the (101)
plane as the predominant plane of crystallization has been iden-
tified as the main phase (Fig. 4c). The anatase structure was
also obtained on a silicon substrate at a deposition tempera-
ture of 500 ◦C [48]. However, minor phases of rutile may be
present, which could not be identified due to the thin (50-nm)
layer.
The growth of zeolite Beta coatings at 150 ◦C is remarkably
enhanced by the superhydrophilicity of the TAMo substrate.
A TAMo substrate without UV irradiation has a coverage of
ca. 50% of a single crystal layer (Fig. 5a), whereas the UV-
irradiated substrate was completely covered with a single layer
of zeolite Beta crystals after a synthesis time of 16 h (Fig. 5b).
The increased hydrophilicity of the substrate resulted in a bet-
ter compatibility with the precursor gel. The zeolite crystals in
the coating evolved from an amorphous gel layer, which was
initially attached to the surface of the substrate [47]. The gel
phase is still partly visible in Fig. 5b. Zeolite Beta crystals were
firmly attached to the substrate due to the increased number
of defect sites and hydroxyl groups on the TiO2 surface. This
clearly demonstrates that the wettability of the surface of the
TAMo substrates influenced the zeolite Beta coating properties.
Similar findings were reported in a study on the improvement
of zeolite NaA nucleation sites on (001) rutile by means of UV
radiation [34].
Fig. 6. Zeolite Beta coverage and average layer thickness of the coatings (a.s.)
◦
obtained on substrate S as a function of synthesis time at 140 and 150 C.
4
◦
−1
Heating rate: 10 C min . The shaded area indicates the coverage of a sin-
gle layer of zeolite Beta crystals. Also the coverage after a repeated synthesis
(ꢀ, 2 × 48 h) is shown.
3.3. In situ synthesis of a single layer of zeolite Beta crystals
The crystallization of zeolite Beta takes from 20 h to several
weeks, depending on the synthesis conditions [47,50–52]. Zeo-
lite Beta was obtained after 20 h from a Na-, K-, and TEAOH-
containing precursor gel with a SiO2/Al2O3 ratio of 50 and
a temperature of 135 ◦C [51]. Therefore, this recipe was taken
as a good starting point for further optimization. In this study,
the synthesis temperature was increased to 140 or 150 ◦C, and
silica sol was used as a reactive silica source to obtain higher
nucleation and crystallization rates.
A single layer of zeolite Beta crystals on substrate S4 cor-
responds to a zeolite coverage between 1.0 and 1.3 g m−2 de-
pending on the average crystal size. A single layer of zeolite
Beta crystals was obtained after 48 and 26 h at 140 and 150 ◦C,
respectively (Fig. 6). Correspondingly, the induction period de-
creased from 30 to 24 h. Increasing the reaction temperature
from 140 to 150 ◦C caused an increase in the crystallization
rate, along with a simultaneous increase in the average crystal
size from 0.82 (Fig. 7a) to 1.57 µm (Fig. 7b), while a narrow
unimodal crystal size distribution (CSD) was preserved. The
narrow CSD indicates that the crystals were grown from a gel
layer initially deposited at the substrate surface. Initially, all

M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
335
Figs. 8a and 8c show SEM images of the zeolite Beta coat-
ings obtained at 140 and 150 ◦C after 48 h and 30 h, respec-
tively. The SEM inset (Fig. 8a) and the AFM scan (Fig. 8b)
show the perfectly closed single layer of well-arranged zeolite
Beta crystals on the TAMo substrate at 140 ◦C. A time inter-
val of just 2 h (between 25 and 27 h of synthesis) corresponds
to a single layer of crystals (1–1.3 g m−2) at 150 ◦C (Fig. 6).
After 30 h, some crystals were already present on top of the
coating (Fig. 8c). Moreover, after rapid heating (10 ◦C min−1
)
of the synthesis mixture at 150 ◦C, the zeolite Beta coating con-
sisted of two layers [48]. The inner layer, comprising relatively
small crystals, was grown directly on the surface at the sub-
strate/gel interface, whereas the visible outer layer of larger
crystals was crystallized into the gel layer starting from the gel–
liquid interface. The AFM scan (Fig. 8d) shows a rougher coat-
ing morphology compared with the coating obtained at 140 ◦C
(Fig. 8b). The crystal shape seems independent on the reac-
tion temperature. At both temperatures, the crystals do not have
a well-defined crystal habit, but appear round-shaped. Some
sharp edges or terraces can be observed on the large zeolite Beta
crystals obtained at 150 ◦C. The zeolite Beta coating thickness
could be increased by up to approximately 4 µm by repeating
the synthesis procedure at 150 ◦C (Fig. 6). A continuous coat-
ing with a rough morphology was obtained (Fig. 9a). The SEM
image in Fig. 9b shows the cross-section of the coating.
Zeolite Beta was the only crystalline phase present in the
coatings. Some selected XRD results of the coatings are pre-
sented in Figs. 10d–10f. The XRD patterns correspond well
with the reference patterns of as-synthesized and calcined ze-
olite Beta powder (Figs. 10a and 10b) obtained from a similar
precursor gel composition as used in this study [43]. Crystal-
lographic faulting is often observed in the zeolite Beta struc-
Fig. 7. Crystal size distribution of zeolite Beta coatings obtained on substrate
◦ ◦
S
4
at (a) 140 C after 48 h, and (b) 150 C after 30 h corresponding to the SEM
images of Figs. 9a and 9c. The synthesis conditions are the same as those in
Fig. 6.
crystals grew fast due to the very high concentration of nutrients
in both the gel layer and the liquid phase. The crystal growth
rate reduced rapidly as the concentrations of nutrients were de-
pleted in the gel layer in the course of crystallization. Therefore,
zeolite crystals with the same size at the end of the crystal-
lization could be formed at different nucleation times and pass
different crystal growth periods [47]. The zeolite Beta crystal
size can be further decreased by a reduction in both the tem-
perature of crystallization [50] and the concentration of alkali
cations, especially potassium [53].
2
Fig. 8. Scanning electron micrographs (a and c) and the corresponding AFM scans (scan area is 10 × 10 µm ) (b and d) of zeolite Beta coatings (a.s.) obtained on
◦
◦
substrate S at 140 C after 48 h (a and b), and 150 C after 30 h (b and d). Close-up SEM images of the coatings are shown in the insets (scale bar indicates 10 µm).
4
The synthesis conditions are the same as those in Fig. 6.

336
M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
Fig. 9. Scanning electron micrographs of (a) the surface and (b, tilting angle =
◦
75 ) the cross section of the zeolite Beta coating obtained after a repeated syn-
◦
thesis at 150 C (see also Fig. 6). The layer thickness of the coating is 4 µm.
The synthesis conditions are the same as those in Fig. 6.
Fig. 11. Activity and product selectivity of the Co-Beta coating (Si/Al = 21,
Co/Al = 0.5) as a function of the reaction temperature. Product Y denotes a C
1
containing product, most likely HCN. The coatings were tested at a total flow
−1
rate of 33 mL min (STP) consisting of 2 vol% of ethylene, 2 vol% ammonia
and 2 vol% oxygen, He-balance.
3.4. Activity of Co-Beta coatings in ethylene ammoxidation
Co-Beta coatings were prepared from the as-synthesized
coatings with a Si/Al ratio of 23 through a series of calcination
and ion-exchange treatments (Table 1). A Co/Al ratio of 0.5
was obtained in the coatings after ion exchange with 10−4
M
aqueous cobalt acetate solution. The Co-zeolite Beta coatings
were tested in a microstructured reactor in a highly exothermic
ethylene ammoxidation. A maximum rate of acetonitrile for-
mation of 220 µmol g−1 s−1 was reached at 500 ◦C (Fig. 11).
Bare Co2+ cations, balanced by two adjacent aluminum atoms
in the zeolite framework, are the most active sites in the am-
moxidation process [57,58]. Thus, higher rates of reaction can
be obtained by increasing the aluminium content in the zeolite
framework. At a Si/Al ratio of 23, the number of aluminium
pairs in the zeolite Beta topology was 30% of the total num-
ber of aluminium atoms; at a Si/Al ratio of 13, this value was
80% [59,60].
The highest carbon-selectivity (SC) to acetonitrile was only
42%, which is two times lower than that reported over Co-Beta
pelletized catalysts with a Si/Al ratio of 12 [62]. Furthermore,
methanol was formed on the coatings with a selectivity of 20%
at 500 ◦C but was not a product on pelletized Co-Beta catalysts.
In the presence of water, methanol can be formed on two adja-
cent bivalent cations [63],
Fig. 10. Reference XRD patterns of (a) as synthesized zeolite Beta powder [54]
and (b) calcined zeolite Beta powder [54], and (c) of the TAMo substrate. XRD
◦
patterns of zeolite Beta coatings: (d) as synthesized, 150 C, 48 h, (e) calcined,
◦
◦
150 C, 48 h, and (f) repeated synthesis, as synthesized, 140 C, 48 h.
ture [2]. The broad peaks from both the (100) and (101) reflec-
tions in the 6–8◦ 2θ range indicates that polymorphs A and B
are approximately equally distributed in the BEA structure [55,
56]. Along with the (100) and (101) reflections, the XRD pat-
terns from the coatings contain strong reflections, with (hkl)
values equal to (205), (302), (304), and (306). The zeolite Beta
crystals in the coating have a preferred orientation, because the
reflections from the crystallographic planes parallel to the crys-
tallographic c-axis are not present in the XRD pattern (cf. the
XRD patterns in Figs. 10a and 10d of as-synthesized zeolite
Beta powder and coating, respectively), whereas the (201) re-
flection is clearly visible (cf. the XRD patterns in Figs. 10b
and 10e of calcined zeolite Beta powder and coating, respec-
tively). A repeated synthesis at 140 ◦C resulted in a multilayer
of zeolite Beta crystals. Fig. 10f shows the XRD pattern of this
as-synthesized coating, in which the (008) reflection is visi-
ble. Apparently, the preferred orientation of crystals disappears
when they grow or are deposited on top of the first layer of crys-
tals on the TAMo substrate.
CH3CN + H2O → CH3OH + HCN.
A constant N-selectivity (SN) to nitrogen of 80% was obtained,
confirming the selective oxidation of ammonia under the ap-
plied reaction conditions. Recently, we showed that higher re-
action rates indeed were obtained over Co-Beta coatings with
a Si/Al ratio of 15 compared with Co-Beta pelletized catalysts,
due to the absence of transport limitations in the coatings [61].

M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
337
obtained when the aluminum content in the zeolite Beta coat-
ing is increased.
Acknowledgments
This work was supported by the Dutch Technology Founda-
tion (STW, project no. EPC.5543), Shell International Chem-
icals B.V., Akzo Nobel Chemicals B.V., and Avantium Tech-
nologies B.V. The authors thank Dr. Milja Mäkelä of Nanoscale
Oy (Finland) for assistance with the development of the ALD
procedure, and Dr. Denis Ovchinnikov for assistance with the
atomic force microscopy.
References
[1] M. Stöcker, Microporous Mesoporous Mater. 82 (2005) 257.
[2] J.C. Jansen, E.J. Creyghton, S.L. Njo, H. van Koningsveld, H. van
Bekkum, Catal. Today 38 (1997) 205.
[3] D.E. de Vos, P.A. Jacobs, Microporous Mesoporous Mater. 82 (2005) 293.
[4] A. de Lucas, M.J. Ramos, F. Dorado, P. Sánchez, J.L. Valverde, Appl.
Catal. A 289 (2005) 205.
[5] Z.-W. Liu, X. Li, K. Asami, K. Fujimoto, Appl. Catal. A 300 (2006) 162.
[6] H.-H. Chen, S.-C. Shen, X. Chen, S. Kawi, Appl. Catal. B 50 (2004) 37.
[7] A.P. Ferreira, C. Henriques, M.F. Ribeiro, F.R. Ribeiro, Catal. Today 107–
108 (2005) 181.
[8] E. Creyghton, Ph.D. thesis, Delft University of Technology, 1996.
[9] M. Lashdaf, V.-V. Nieminen, M. Tiitta, T. Venäläinen, H. Österholm,
O. Krause, Microporous Mesoporous Mater. 75 (2004) 149.
[10] Y. Li, J.N. Armor, J. Catal. 173 (1998) 511.
Fig. 12. Method for the in situ synthesis of zeolitic coatings. Surface modi-
fication of the molybdenum (Mo) substrate: (1) cleaning and etching of the
substrate, (2) deposition of a 170 nm Al O film by atomic layer deposition
2
3
(ALD), (3) deposition of a 50 nm TiO film by ALD. Step (4) illustrates the
2
in situ crystallization process of zeolitic coatings on the TAMo substrate (after
a UV treatment of the TiO surface).
2
[11] J. Hedlund, O. Öhrman, V. Msimang, E. van Steen, W. Böhringer,
S. Sibya, K. Möller, Chem. Eng. Sci. 59 (2004) 2647.
[12] M. Torres, L. López, J.M. Domínguez, A. Mantilla, G. Ferrat, M. Gutier-
rez, M. Maubert, Chem. Eng. J. 92 (2003) 1.
4. Conclusion
A method has been developed for the in situ synthesis of
zeolitic coatings on a metal substrate for application in a mi-
crostructured reactor. The method is exemplified in Fig. 12 for
the in situ growth of a single layer of zeolite Beta crystals on
a molybdenum substrate. First, the number of surface irregu-
larities on the substrate was increased by a chemical etching
procedure (step 1) to enhance both the number of nucleation
sites and the positions of zeolite crystals for anchoring. For
a single layer of zeolite Beta crystals 0.5–1.5 µm in size, a low
surface roughness (14 nm) was required. Then, two thin films of
Al2O3 (170 nm, step 2) and TiO2 (50 nm, step 3) were succes-
sively deposited by ALD on the substrate. The internal Al2O3
film protects the Mo substrate from oxidation up to 550 ◦C
in an oxidative environment. High wettability of the external
TiO2 film after UV irradiation increases zeolite nucleation at
its surface. A stable superhydrophilic TiO2 surface (water con-
tact angle below 5◦) for at least 200 min was obtained after
deposition of a 50 nm film at 500 ◦C from a TiCl4 precursor.
A perfectly ordered single layer of zeolite Beta crystals was ob-
tained from a viscous Na, K, and TEAOH precursor gel with
a SiO2/Al2O3 ratio of 46 at a synthesis temperature of 140 ◦C
after 48 h (step 4).
[13] O.L. Oudshoorn, M. Janissen, W.E.J. van Kooten, J.C. Jansen, H. van
Bekkum, C.M. van den Bleek, H.P.A. Calis, Chem. Eng. J. 54 (1999) 1413.
[14] N. van der Puil, F.M. Dautzenberg, H. van Bekkum, J.C. Jansen, Micro-
porous Mesoporous. Mater. 27 (1999) 95.
[15] L. Li, B. Xue, J. Chen, N. Guan, F. Zhang, D. Liu, H. Feng, Appl. Catal.
A 292 (2005) 312.
[16] A.E.W. Beers, T.A. Nijhuis, N. Aalders, F. Kapteijn, J.A. Moulijn, Appl.
Catal. A 243 (2003) 237.
[17] A. Bueno-López, D. Lozano-Castelló, I. Such-Basáñez, J.M. García-
Cortés, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea, Appl. Catal. B 58
(2005) 1.
[18] E.V. Rebrov, G.B.F. Seijger, H.P.A. Calis, M.H.J.M. de Croon, C.M. van
den Bleek, J.C. Schouten, Appl. Catal. A 206 (2001) 125.
[19] J. Coronas, J. Santamaria, Chem. Eng. Sci. 59 (2004) 4879.
[20] J.C.M. Muller, G. Hakvoort, J.C. Jansen, J. Therm. Anal. 53 (1998) 449.
[21] Y. Shan, S. Wan, J.L.H. Chau, A. Gavriilidis, K.L. Yeung, Microporous
Mesoporous Mater. 42 (2001) 157.
[22] M.J.M. Mies, E.V. Rebrov, M.H.J.M. de Croon, J.C. Schouten, Chem.
Eng. J. 101 (2004) 225.
[23] M.J.M. Mies, J.L.P. van den Bosch, E.V. Rebrov, J.C. Jansen, M.H.J.M.
de Croon, J.C. Schouten, Catal. Today 110 (2005) 38.
[24] E.V. Rebrov, M.H.J.M. de Croon, J.C. Schouten, Book of Abstracts of 7th
European Conference on Catalysis, EuropaCat VII, August 28–September
1, 2005, Sofia, Bulgaria (2005), p. 321.
[25] S.A. Kuznetsov, S.V. Kuznetsova, E.V. Rebrov, M.J.M. Mies, M.J.H.M.
de Croon, J.C. Schouten, Surf. Coat. Technol. 195 (2005) 182.
[26] S.A. Kuznetsov, E.V. Rebrov, M.J.M. Mies, M.H.J.M. de Croon, J.C.
Schouten, Surf. Coat. Technol. 201 (2006) 971.
[27] M.D. Groner, J.W. Elam, F.H. Fabreguette, S.M. George, Thin Solid
Films 413 (2002) 186.
[28] N.D. Hoivik, J.W. Elam, R.J. Linderman, V.M. Bright, S.M. George, Y.C.
Lee, Sens. Actuators A 103 (2003) 100.
The Co-Beta coatings (Si/Al = 23, Co/Al = 0.5) were
tested in the ammoxidation of ethylene to acetonitrile in a mi-
crostructured reactor. A maximum rate of reaction of 220 µmol
C2H3N g−1 s−1 was obtained at 500 ◦C, with a carbon selec-
tivity to acetonitrile of 42%. Higher rates of reaction can be

338
M.J.M. Mies et al. / Journal of Catalysis 247 (2007) 328–338
[29] J.C. Jansen, W. Nugroho, H. van Bekkum, in: R. Von Ballmoos, J.B. Hig-
gins, M.M.J. Treacy (Eds.), Proceedings of the 9th International Zeolite
Conference, Montreal, 5–10 July 1992, p. 247.
[30] Z. Wang, Y. Yan, Microporous Mesoporous Mater. 48 (2001) 229.
[31] V. Valtchev, S. Mintova, L. Konstantinov, Zeolites 15 (1995) 679.
[32] V. Valtchev, S. Mintova, Zeolites 15 (1995) 171.
[33] J.C. Jansen, J.H. Koegler, H. van Bekkum, H.P.A. Calis, C.M. van den
Bleek, F. Kapteijn, J.A. Moulijn, E.R. Geus, N. Van der Puil, Microporous
Mesoporous Mater. 21 (1998) 213.
[46] M. Schuisky, K. Kukli, J. Aarik, J. Lu, A. Hårsta, J. Cryst. Growth 235
(2002) 293.
[47] M.J.M. Mies, E.V. Rebrov, J.C. Jansen, M.H.J.M. de Croon, J.C.
Schouten, Microporous Mesoporous Mater. (2007), http://dx.doi.org/
10.1016/j.micromeso.2007.02.032.
[48] M. Ritala, M. Leskelä, E. Nykänen, P. Soininen, L. Niinistö, Thin Solid
Films 225 (1993) 288.
[49] G. Krautheim, T. Hecht, S. Jakschik, U. Schröder, W. Zahn, Appl. Surf.
Sci. 252 (2005) 200.
[50] J. Pérez-Pariente, J.A. Martens, P.A. Jacobs, Zeolites 8 (1988) 46.
[51] M.A. Camblor, A. Mifsud, J. Pérez-Pariente, Zeolites 11 (1991) 792.
[52] R.L. Wadlinger, G.T. Kerr, E.J. Rosinski, U.S. Patent 3,308,069 (1967).
[53] M.A. Camblor, J. Pérez-Pariente, Zeolites 11 (1991) 202.
[54] H. Robson, Microporous Mater. 22 (1998) 551.
[55] M.M.J. Treacy, J.B. Higgins, R. von Ballmoos, Zeolites 16 (1996)
323.
[34] A.W.C. van den Berg, L. Gora, J.C. Jansen, T. Maschmeyer, Microporous
Mesoporous Mater. 66 (2003) 303.
[35] L. Sirghi, Y. Hatanaka, Surf. Sci. 530 (2003) 323.
[36] X. Wang, Y. Yu, X. Hu, L. Gao, Thin Solid Films 371 (2000) 148.
[37] J. Aarik, A. Aidla, T. Uustare, V. Sammelselg, J. Cryst. Growth 148 (1995)
268.
[38] J. Aarik, A. Aidla, V. Sammelselg, T. Uustare, J. Cryst. Growth 181 (1997)
259.
[56] B.J. Schoeman, E. Babouchkina, S. Mintova, V.P. Valtchev, J. Sterte,
J. Porous Mater. 8 (2001) 13.
[39] J. Aarik, A. Aidla, H. Mändar, T. Uustare, Appl. Surf. Sci. 172 (2001) 148.
[40] D.R.G. Mitchell, D.J. Attard, G. Triani, Thin Solid Films 441 (2003) 85.
[41] D.R.G. Mitchell, G. Triani, D.J. Attard, K.S. Finnie, P.J. Evans, C.J. Barbé,
J.R. Bartlett, Device and Process Technologies for MEMS, Microelectron-
ics and Photonics III, in: J.C. Chiao, A.J. Hariz, D.N. Jamieson, G. Parish,
V.K. Varadan (Eds.), in: Proceedings of the SPIE, vol. 5276, 2004, p. 296.
[42] M. Schuisky, A. Hårsta, A. Aidla, K. Kukli, A.-A. Kiisler, J. Aarik, J. Elec-
trochem. Soc. 147 (2000) 3319.
[57] R. Bulánek, K. Novoveská, B. Wichterlová, Appl. Catal. A 235 (2002)
181.
[58] J. Deˇdecˇek, D. Kaucý, B. Wichterlová, Microporous Mesoporous Mater.
35–36 (2000) 483.
[59] O. Bortnovsky, Z. Sobalík, B. Wichterlová, Microporous Mesoporous
Mater. 46 (2001) 265.
ˇ
[60] B. Wichterlová, J. Deˇdecˇek, Z. Sobalík, J. Cejka, Stud. Surf. Sci. Catal.
135 (2001).
[43] D.R.G. Mitchell, D.J. Attard, G. Triani, J. Cryst. Growth 285 (2005) 208.
[44] K. Kukli, A. Aidla, J. Aarik, M. Schuisky, A. Hårsta, M. Ritala, M. Les-
kelä, Langmuir 16 (2000) 8122.
[61] M.J.M. Mies, E.V. Rebrov, C.J.B.U. Schiepers, M.H.J.M. de Croon,
J.C. Schouten, Chem. Eng. J. Chem. Eng. Sci. (2007), http://dx.doi.org/
10.1016/j.ces.2006.12.018.
[45] M. Schuisky, J. Aarik, K. Kukli, A. Aidla, A. Hårsta, Langmuir 17 (2001)
5508.
[62] Y. Li, J.N. Armor, J. Catal. 173 (1998) 511.
[63] T. Lu, X. Zhuang, Y. Li, S. Chen, J. Am. Chem. Soc. 126 (2004) 4760.