A novel method to prepare macroporous TiO2 ceramic, based on membrane emulsification was reported. To solve the paradox between the instability of nonaqueous emulsion and long emulsification time required by the membrane emulsification, a two-stage ceramic membrane jet-flow emulsification was proposed. Discussion was conducted on the evolution of droplet size with time, which followed the Ostwald ripening theory. And a monodispersed nonaqueous emulsion with an average droplet size of 1.6μm could be prepared. Using the emulsion as a template, TiO2 ceramics with an average pore size of 1.1μ.m were obtained. The material could be prospectively used for preparation of catalysts, adsorbents, and membranes.
Porous materials have recently received much
attention because of their application in a wide variety
of fields, such as biosensors, catalysts, adsorbents,
chromatographic materials, and photonic crystals.
Several approaches are currently available for the
preparation of ordered structures with different length
scales. Mesostructures can be obtained by the
self-organization of the surfactant molecules, and colloidal
templates are usually used to prepare ordered
macroporous materials[ 1,2]. Imhof and Pine[3] have
proposed a famous method - the emulsion templating
approach, for preparing highly ordered macroporous
materials. In this approach, the sol-gel process is
used to deposit an inorganic material on the exterior of
the droplets in a nonaqueous emulsion. The preparation
of monodispersed nonaqueous emulsion is one of
the key processes.
Membrane emulsification (ME) is a new emulsification
technique based on membrane structure,
which has received increasing attention over the last
10 years, because it consumes less energy, has low
shearing stress and controllable droplet size of the
emulsions[4]. However, low flux of the dispersed
phase, resulting in a long emulsification time, has
hindered the application of this technology. Especially
in the preparation of nonaqueous emulsions, the problem
has become more obvious because of nonaqueous
emulsions being unstable.
In the previous study[5], a membrane jet-flow
emulsification to resolve the paradox between the flux
and the droplet size, in conventional membrane emulsification,
was proposed. A monodispersed OW
emulsion could be obtained under jet flow by a
two-stage membrane jet-flow emulsification process.
The main purpose of this study was to prepare amonodispersed nonaqueous emulsion with the help of
membrane emulsification under jet-flow, to investigate
the stability of the nonaqueous emulsion, and to prepare
an ordered macroporous Ti02 ceramic based on
this method.
In these experiments, the stable emulsions were
prepared with formamide as the continuous phase,
isooctane as the dispersed phase, an emulsifier concentrate
of 2%, and a Zr02 membrane with nominal
size 0.16y.m was used in the first stage. Fig.1 shows
the emulsion prepared at a pressure of 0.09MPa, and
the rotate speed of 300rmin-'. As shown in this figure,
the average droplet size of the emulsion was 3.2ym
and the a was 0.563. The emulsion prepared was not
monodispersed because of the higher pressure, however'the
flux could reach 201.4LK2.h-'. Using the
secondary jet-flow membrane emulsification, monodispersed
emulsion with small droplet size could be
obtained. Fig.2 shows the droplet size of the emulsion
prepared by using an a-A1203 ceramic membrane,
having a nominal pore size of 1.5ym, at 15OkPa pressure,
with 300rmin - I stirring speed. The average
droplet size of the emulsion was 1.6ym, which was
slightly bigger than the membrane pore size. The a of
the emulsion was 0.22, which meant the emulsion was
monodispersed. The flux reached 176.4L.m-2.hp I ,
which was attributed to the larger pore size and higherpressure. The stability of the emulsion is shown in
Fig.3. The cube of average droplet diameter of the
emulsion linearly increased with time, which followed
the famous Ostwald ripening theory[8]. And the coefficient
of dispersion (a) had little changes from 0.22 to
0.24, which meant the emulsion had always been kept
monodispersed during the 960min. Therefore, the stability
of the emulsion could meet the demands of the
sol-gel process.A light-scattering particle sizer (Zetasizer 3000,
Malvern Instruments, Malvern, UK), determined the
particle size of the sol. The evolution of the sol with
time is shown in Fig.4. The particle size of the sol was
about 4nm initially, and it became more than lOnm
after 10 days. The result indicated that the particle size
increased with time because of coalescence. Generally
the sol was stable and no precipitation emerged in
several weeks, however, it was recommended to use
the sol in three days to obtain the ordered porous material.The emulsion was concentrated by centrifuging
and then dispersed in the sol, and the stable gel could
be obtained by adding 30% ammonia into the mixture.
The template was removed after heating to a temperature
of 40°C for 24h, and then the gel was calcined at
a temperature of 400°C in a furnace, for 4h, to obtain
macroporous ceramic. Fig.5 shows the scanning electron
micrographs of the surface and cross-section. The
Ti02 ceramic with the pore size of 1.1pm was obtained
by the calcining process, which was about 0.69times of the droplet size of the emulsion before ther- 2
mal treatment.
In this study the droplet size of the emulsion template
was controlled by the pore structure of the memtemplate
determined the macroporous structure of the
Ti02 ceramic. Therefore, it was possible to control the
macroporous structure of the material by choosing the
membranes with different pore structures. Further study
on the control of porous structure is still in progress.