Hollow spheres
Martin Sgraja1
Jan Blömer1
Research Article
Jürgen Bertling1
Peter J. Jansens2
Thermal and Structural Characterization of
TiO2 and TiO2/Polymer Micro Hollow
Spheres
1
2
Fraunhofer Institute for
Environmental, Safety, and
Energy Technology UMSICHT
– Special Materials,
Oberhausen, Germany.
Delft University of Technology,
Department of Process &
Energy Laboratory, Delft, The
Netherlands.
2029
Micro hollow spheres, synthesized by the coating of tetradecane filled microcapsules with titanium dioxide, are characterized using thermal gravimetry (TG),
infrared spectroscopy (IR), Hg-porosimetry and scanning electron microscopy
(SEM). The investigations show a strong dependence of the coating efficiency on
the initial pH (precipitation in water or dilute sulfuric acid) as well as the specific
capsule surface present in solution. Since the process is dominated by heterogeneous precipitation, the coating efficiency is governed by the counteracting
processes of capsule-TiO2 and TiO2-TiO2 agglomeration. TG-IR analysis of the
capsules shows the vaporization of tetradecane prior to the decomposition of the
polymeric wall to carbon monoxide, carbon dioxide and water. After the extraction or calcination of the core microcapsules, stable inorganic and organic-inorganic hollow spheres are obtained.
Keywords: Hollow spheres, Microcapsules, Thermal hydrolysis
Received: March 20, 2010; revised: July 6, 2010; accepted: July 27, 2010
DOI: 10.1002/ceat.201000288
1
Introduction
Micrometer hollow spheres have a wide range of applications.
The most common of these is probably their application as fillers for weight reduction and the associated reduction in costs.
In this case, spherical hollow spheres show lower specific gravities, better flowability and higher solid contents compared to
irregular particles due to their low surface to volume ratios.
Other applications for hollow spheres include their use as
photocatalysts, gas or chemical storage devices and reactors as
well as drug delivery systems.
Spray processes and template sacrificial methods can be used
for the production of such hollow spheres. In spray processes
such as spray drying and spray pyrolysis, wall material solutions are sprayed into a hot gas forming hollow spheres due to
the vaporization of the liquid and crystallization or precipitation of the solid [1–3]. In general, although these processes
allow high throughputs at low costs, they are known to suffer
from a lack of distinct process control, e.g., with respect to the
size distribution of the products. In template sacrificial methods, a template is coated in a solution of the desired wall material, and subsequently, the template is removed by dissolution,
vaporization or decomposition, leaving back a hollow sphere
of the wall material.
–
Correspondence: Mr. M. Sgraja (sgraja@muenster.de), Helenenstraße 101, 41748 Viersen, Germany.
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
Apart from the differentiation according to the wall formation process, a division into processes with soft and hard templates is possible. Soft templates include emulsion droplets
and vesicles whereas hard templates are organic or inorganic
particles. Emulsion droplets as templates are used in such
applications as the emulsion extraction method [4] and the
hydrolyzation of alkoxides from the inside of oil-in-water
emulsions [5]. In template sacrificial methods, different hard
template coating processes such as layer-by-layer techniques,
sol-gel methods, liquid phase deposition, and forced hydrolysis
have all been investigated [6–11]. The advantages of liquid
processes are the distinct process control, e.g., predefinition of
the size distribution through the selection of the template
sizes, as well as the large number of different core and coating
materials that can be used to formulate hollow spheres with a
layered structure composed of different materials.
In the present paper, the template sacrificial route was applied using amino-resin microcapsules as the core and titanium dioxide as the model wall material. In contrast to solid
core particles, liquid filled microcapsules have the advantage
that organic-inorganic hybrid capsules can be formed after the
coating step when the liquid from the inside of the capsules is
removed. The capsules used were synthesized by an in situ
polymerization of melamine-formaldehyde pre-condensed
from an aqueous phase. For this purpose, tetradecane was chosen as an example core material and was emulsified in an aqueous phase followed by dissolution of the precondensates in the
continuous phase. Subsequently, the wall formation, i.e., con-
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cet-journal.com
2030
M. Sgraja et al.
densation of the prepolymer at the surface is initiated at elevated temperature by decreasing the pH with citric acid. After
the hardening of the capsule wall, the capsules can be separated from the continuous phase by filtration and can subsequently be used for the coating process. In a previous paper, it
was shown that the size of the hollow spheres can be controlled
by the size of the capsules, which is again controlled by the size
of the emulsion droplets at the beginning of the encapsulation
[12]. In addition, it was found that the surface averaged mean
diameter of the emulsion droplets can be calculated based on
the theory of Kolmogorov using the interfacial tension, the
stirring speed, the stirrer diameter, the fluid density and the
phase volume ratio of the dispersed and continuous phase
[12]. While the thickness of the polymeric wall as well as the
size of the capsules can be controlled in the encapsulation process, in the coating process the properties of the outer wall,
e.g., thickness, are adjusted.
In the present work, titanium dioxide precipitated by thermal hydrolysis of titanium oxysulfate is used for the coating
process. This process is based on the hydrolyzation of the titanium salt at elevated temperatures, condensation of the hydrolyzates up to a certain size where nucleation occurs, and subsequent growth and agglomeration of the particles. The coating
is performed in solutions of different initial pH, i.e., in water
and 0.1 M H2SO4 solutions and the coated capsules and the
hollow spheres are characterized by thermal gravimetry and
thermal gravimetry coupled with infrared spectroscopy. In
addition, mercury porosimetry measurements are performed
to determine the stability and the porosity of the capsules and
scanning electron microscopy (SEM) is used to examine the
surface structure of the products.
2
Experimental
2.1
Materials
Poly(ethylene glycol) with an average molecular weight of
400 g L–1 was purchased from Fluka Chemical Co. (Baus, Switzerland). Citric acid, anhydrous for synthesis, was purchased
from Merck Schuchardt Co. (Hohenbrunn, Germany). Methanol etherified melamine-formaldehyde precondensates with a
solids content of 87 wt-% were provided by Cytec Industries
BV (Botlek-Rotterdamm, The Netherlands). Tetradecane (Linpar 14) was purchased from Condea (Duisburg, Germany).
Titanium(IV) oxysulfate was purchased from Sigma-Aldrich
(Seelze, Germany). 0.1 M H2SO4 was prepared by dilution of
concentrated sulfuric acid purchased from Sigma-Aldrich
(Seelze, Germany). Separation of the precipitates from solutions was either carried out using a polypropylene (PP) membrane Accurel PP 2E HF R/P (Membrana, Germany) with a
nominal pore size of 0.2 lm or by syringe filters Rotilabo CME
(Carl Roth GmbH, Germany) with a pore size of 0.22 lm.
2.2 Thermal and Structural Characterization
Thermal gravimetry (TG) and thermal gravimetry coupled
with infrared spectroscopy (TG-IR) were used for the charac-
www.cet-journal.com
terization of the thermal properties of the resin and titanium
dioxide as well as the coated and uncoated capsules. The latter
technique allows the direct identification of the chemical composition of the gaseous products occurring during the thermal
treatment. TG and TG-IR analysis were performed using a
STA409 (Netzsch, Germany) or a TG 209 F1 Iris (Netzsch, Germany) in combination with a spectrometer Tensor (Bruker
Optics, Germany). In all the measurements, a heating rate of
10 K min–1 and a N2/O2-atmosphere (80 vol.-%/20 vol.-%)
were applied.
Mercury porosimetry measurements were used (Pascal 140
and Pascal 440, Porotec inc., Hofheim) for the characterization
of the pore size distribution and the stability of the capsule. In
addition scanning electron microscopy (ESEM Quanta 400,
Fei Company) was used to characterize the surface of the coated and uncoated capsules. Prior to SEM investigations, the
samples were coated with a thin conducting layer of gold/palladium (20 %) (Sputter Coater K550, Emitech).
2.3
Microencapsulation
Microcapsules were synthesized by an in situ polymerization
using melamine-formaldehyde precondensates. A detailed description of the process is given elsewhere [12]. Typically, 1.2 g
poly(ethylene glycol) acting as a protective colloid was dissolved in 80 mL of distilled water and heated up to 60 °C. In
parallel, 40 mL of tetradecane as the core material was heated
up to 60 °C. Subsequently, the tetradecane was emulsified in
the aqueous phase at the same temperature using a Polytron
1300D system (Kinematica AG, Switzerland). Meanwhile, 2.4 g
melamine-formaldehyde precondensates was dissolved in
water giving a total volume of 20 mL. The wall-material solution was added to the poly(ethylene glycol) solution following
10 min of emulsification and additional emulsification was
allowed for a further 10 min. Shortly before the addition, the
pH of the wall material solution was adjusted to 3 to initiate
the condensation reaction. After the emulsification, the capsules were hardened while stirring the solution with a magnetic
stirrer (300 s–1) at 60 °C for 2 h. For the characterization of the
capsules, the solution was filtered, washed twice with ca.
100 mL distilled water and dried under ambient conditions or
stored under wet conditions for thermal analysis.
2.4
Determination of the Weight Percentages
of the Capsule Wall
Density measurements using a gas pycnometer (Pycnomatic
ATC, Porotec inc.) with helium as the measurement gas were
used for the determination of the mass fraction of the capsule
wall. Afterwards, the weight percentage of the capsule wall,
wresin, was derived from the density of the capsules, qcapsule,
with the value lying between the density of pure tetradecane
(qC14H30 = 0.76 g cm–3 [13] and that of the pure precipitated
resin (qresin = 1.56 g cm–3), Eq. (1). With the help of Eq. (2),
the weight percentage of tetradecane wC14H30 was substituted
in Eq. (1) and yields Eq. (3) following some rearrangement:
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
Hollow spheres
wC H
1
w
resin 14 30
qcapsule qresin qC14 H30
(1)
wresin wC14 H30 1
(2)
qC14 H30 qresin
wresin
2.5
qcapsule
qC14 H30
qresin
(3)
qresin
Coating with Titanium Dioxide
For the precipitation of the titanium dioxide, titanium(IV)
oxysulfate was dissolved in 60 mL distilled water and the clear
solution was filtered through a 0.2 lm polypropylene membrane to remove particles and impurities. In parallel, 452 mL
of distilled water, including the appropriate amount of sulfuric
acid and the microcapsules, were heated up to 84 °C while stirring at 300 s–1. On reaching a constant temperature, 48 mL of
the TiOSO4-solution were added under continuous stirring by
means of a syringe pump to give a total volume of 500 mL.
The mass of the titanium oxysulfate was calculated according
by considering an overall concentration of 0.1 mol L–1. By adding the titanium solution of ca. room temperature, the temperature in the vessel drops to the final working temperature of
80 °C. The coating process was complete within 6 h under stirring at 300 s–1 and 80 °C. Afterwards, the solution was centrifuged to separate the capsules from the sedimented titanium
dioxide and the sediment was discarded. The centrifugation
was repeated two to three times until no more settled product
could be observed. The remaining capsules were then centrifuged and washed twice with ca. 50 mL of 0.1 M H2SO4 and
twice with ca. 50 mL distilled water. Subsequently, the capsules
were dried under ambient conditions or stored without drying
for thermal analysis.
After drying of the coated capsules, hollow spheres were
either formed by extracting the tetradecane from within the
microcapsule or by vaporization of the tetradecane and burning of the melamine-formaldehyde wall. In the first method,
the tetradecane was extracted by thermal treatment in a vacuum oven at ca. 120 °C for several days while the melamine-formaldehyde wall and the titanium dioxide layer remained to
form an organic-inorganic hybrid hollow sphere. In the extraction, the temperature was chosen to be low enough to minimize the thermal degradation of the polymer. During the
extraction, the complete loss of the liquid core was checked by
disappearance of the tetradecane bands (2800–3000 cm–1, see
Fig. 3) in the infrared spectra. In the second method, the tetradecane as well as the polymer were removed by calcination of
the products at 800 °C for 4 h leaving only the inorganic hollow spheres of titanium dioxide.
2.6
Determination of the Amount of Titanium
Dioxide on the Capsules
Thermal gravimetry was used for the determination of the
mass percentage of titanium dioxide precipitated on the cap-
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
2031
sules. For this purpose, the capsules were stored under wet
conditions to prevent losses of tetradecane during storage.
Measurement of the samples was undertaken according to the
previously described conditions (see Sect. 2.2). The weight percentage of titanium dioxide was taken from the remaining
mass at 800 °C while the mass percentage was corrected with
respect to the amount of water present in the samples (see Section 3.2). Consequently, the weight percentage of the capsules
(amino resin and tetradecane) can be calculated from the difference of the total mass and the amount of water and titanium dioxide. If required, the weight percentages of tetradecane and amino resin can additionally be determined from
density measurements.
In order to calculate a mean coating thickness s from the
mass percentage of the TiO2 precipitated on the capsules, the
density of the coated capsules was firstly calculated using
the density of titanium dioxide, qTiO2, the mean density of the
capsules, qcap = 0.80 g cm–3, and the weight percentages of
the capsules, wcap, and TiO2, wTiO2, according to Eq. (1).
Consequently, the mean shell thickness can be derived from
Eq. (4):
s 0:5d3;2 1
s!
qproduct
3 qTiO2
qTiO2 qcapsule
(4)
where qproduct corresponds to the coated capsules and d3,2 is
the mean surface weighted capsule diameter determined by
laser diffraction particle size measurements (Mastersizer 2000,
Malvern Instruments, UK) [12].
3
Results and Discussion
3.1 Thermogravimetric Analysis (TGA)
A thermogravimetric analysis of the amino resin was performed initially. The weight loss as a function of the temperature is given in Fig. 1. It can be seen from Fig. 1 that the amino
resin loses its weight in four steps in the temperature range
from room temperature (RT) up to 700 °C. In the first temperature range between RT and 250 °C, the amino resin shows two
steps with an overall weight loss of ca. 25 %, which can be assigned to the loss of water contained in the resin. At ca. 370 °C
there is a sharp decrease of the mass, which can be attributed
to the decomposition of the amino resin. The total mass is
reduced to ca. 50 % at this stage. Subsequently, the mass
decreases continuously until all the material is burned at a
temperature of ca. 700 °C. All of the steps in the decomposition of the amino resin are endothermal processes.
The decomposition of the microcapsules occurs in two
steps, the first between ca. 100–250 °C and the second small
weight loss at ca. 400 °C, Fig. 1. Both processes are endothermal. Due to the large weight loss (≈ 95 %), the first step can be
assigned to the loss of tetradecane, which agrees quite well
with a weight fraction of 95.2 % calculated from the capsule
density (qcapsules = 0.782 g cm–3) measured by gas pycnometry.
It can be seen that the evaporation already starts at ca. 100 °C
and finishes at ca. 250 °C, while the endothermal peak shows
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cet-journal.com
2032
M. Sgraja et al.
ide, carbon dioxide, water and ammonia
are formed before the release of tetradecane. The analysis shows that the release
of gaseous tetradecane (wave number
80
2800–3000 cm–1) already starts at low temperatures below 100 °C while the main
weight loss occurs between ca. 100–200 °C.
60
The peak emanating from the tetradecane
signal is most likely due to the sudden
breakup of capsules (thermal breakup),
40
and thus, an increased liberation of tetradecane. It is worth noting that the ignition
20
temperature of tetradecane at 200 °C (see
[13], 16–30) is covered by this range. However, in Fig. 2, it can be seen that the release
0
of tetradecane is almost finished when CO2
0
100
200
300
400
500
600
700
800
production starts to occur, indicating that
the tetradecane truly evaporates and does
temperature [°C]
not decompose. Thus, the volume concencoated capsules
micro capsules
tration of tetradecane in the gaseous phase
amino resin
titanium dioxide
in this case seems to be below the flammFigure 1. Thermal analysis of amino resin, titanium dioxide, coated and uncoated capability limit of 0.5 % [13].
sules.
After liberation of the liquid core, the
amino resin from the capsule wall starts to
that the tetradecane evaporates without burning. The second
decompose forming carbon dioxide (600–700 cm–1 and 2200–
step, which is coupled to a small weight loss, can consequently
2400 cm–1), a small amount of carbon monoxide (2000–
be assigned to the decomposition of the amino resin.
2200 cm–1), water (1300–1900 cm–1) and ammonia (900–
For a more detailed analysis of the process gases occurring
1000 cm–1), which can hardly be seen due to the low IR-abduring the thermal treatment a TG-IR measurement was persorption.
formed. The three dimensional graph of the dependence of the
The thermal analysis of titanium dioxides (Fig. 1) shows only
spectra on the measurement time and the IR-spec of an overall
two distinct endothermal steps, the first between RT and about
thermal analysis and the corresponding substances in the gas
400 °C and the second in the range of ca. 550–700 °C. The first
phase, are shown in Figs. 2 and 3, respectively.
weight loss (15.3 %) corresponds to the loss of free and interstiIn the three dimensional TG-IR analysis in combination
tial water and the second (8.4 %) to the release of sulfur oxide
with the overall spectrum, it can be seen that carbon monoxliberated by the decomposition of precipitated sulfates [14–16].
Using the weight changes from thermal analysis, the
stoichiometry TiO2(H2O)0.89(SO3)0.11, which is
equal to TiO2(H2O)0.78(H2SO4)0.11 for the precipitated material can be calculated.
In the thermal analysis of the coated capsules
(specific surface of 4 m2L–1 and 20 m2L–1), the
effects of the capsules, i.e., the loss of tetradecane,
is the dominating effect. As seen already in the
decomposition of the capsules, the loss of tetradecane starts at ca. 100 °C and is finished at ca.
250 °C, resulting in a weight loss of ca. 70 %. The
remaining mass at 800 °C can be assigned to pure
titanium dioxide. The effects originating from the
decomposition of the polymeric wall or the loss of
water from titanium dioxide due to the low weight
ratio cannot be seen in the measurement.
rel. mass [%]
100
3.2
Figure 2. TG-IR-analysis of the microcapsules.
www.cet-journal.com
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Determination of the Amount of
Titanium Dioxide on the Capsules
As was described above, thermal analysis of capsules stored under wet conditions was used to
determine the weight percentage of titanium diox-
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
Hollow spheres
Figure 3. Overall IR-spectra of the gaseous products formed during the thermal
analysis of capsules.
ide. In addition to the described effects occurring in the thermal analysis (see Section 3.1), an endothermal weight loss
occurs between RT and 120 °C, Fig. 4. This weight loss due to
the vaporization of water was taken at the step between the
first and the second weight loss (light grey area, Fig. 4). By
comparing the thermal analysis of pure amino resin and that
of the capsules (Fig. 1) it can be seen that the amino resin also
shows a small weight loss in this temperature range. It is reasonable to assume that this is due to the water remaining from
the encapsulation process. Since the amount of the wall material within the capsules (ca. 20 %) and the amount of water in
the resin (ca. 20 %) are quite low, the water originating from
the resin is accounted for by the water from the wet capsules.
The weight percentages of titanium dioxide precipitated on the capsules for the precipitation in
water and diluted sulfuric acid and for specific
capsule surfaces of 4 and 20 m2L–1, are given in
Tab. 1. In addition, the mean shell thicknesses calculated according to Eq. (4) are given in parentheses. Comparing the coating in dilute sulfuric acid
and water at a capsule surface of 4 m2L–1, it can be
seen that the weight percentage of TiO2 precipitated on the capsules in sulfuric acid (24.9 wt %) is
much higher than that in water (14.0 wt. %). When
the acidic TiOSO4 solution is added to water, there
is a sharp increase in the pH before it drops again
when the solution becomes homogeneously mixed.
As a result of this increase in pH, a large number
of particles are formed that can agglomerate directly. However, in addition to the agglomeration
between capsules and TiO2-particles, the agglomeration between TiO2-particles themselves also
takes place. Thus, the TiO2–TiO2 agglomeration
competes with the capsule-TiO2 agglomeration. In
Table 1. Gravimetrically determined weight percentages of TiO2
precipitated in water and sulfuric acid on capsules of a specific
surface of 4 and 20 m2L–1. In addition the corresponding mean
shell thicknesses of the inorganic layer obtained from Eq. (4) are
given in parentheses.
H2SO4
H2O
4m L
24.9 ± 2.0 wt %
(127 nm)
14.0 ± 5.9 wt %
(64 nm)
20 m2L–1
15.3 ± 2.2 wt %
(71 nm)
13.2 ± 2.1 wt %
(60 nm)
2 –1
Figure 4. Thermal analysis of coated capsules for the determination of the
weight percentage of the coating.
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
2033
sulfuric acid, where the particles are formed more
slowly and continuously, the agglomeration between a capsule and a particle is more likely than
the agglomeration between two TiO2-particles,
thus leading to a higher coating efficiency.
In the case of a capsule surface of 20 m2L–1, the
amount of TiO2 on the capsules in water is only
slightly lower (13.2 wt %) than that in sulfuric acid.
This can be explained by the fact that the mean
distance between a capsule and a TiO2-particle decreases with the increasing number of capsules
while the mean distance between the TiO2-particles can be assumed to stay constant. Therefore, in
the beginning when a large amount of TiO2-particles are formed in water, a higher percentage of the
particles are collected by the capsules. It can be
assumed that even with a higher specific surface of
the capsules, i.e., a larger number of capsules, the
difference between the coatings in water and sulfuric acid would vanish.
By comparing the precipitation in sulfuric acid
with specific surfaces of 4 and 20 m2L–1, it can be
seen that the in the latter case, the precipitation of
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cet-journal.com
2034
M. Sgraja et al.
15.3 wt % is considerably lower than for the former surface,
which has a value of 24.9 wt %. This is reasonable since the
percentage of coating material per surface decreases due to an
increased surface and a constant amount of coating material.
The observation that the percentage of coating in H2SO4 at a
surface of 20 m2L–1 is not a fifth of the precipitation at 4 m2L–1
is possibly due to the fact that the underlying process for the
agglomeration is not linear.
From a comparison of the coating thickness of capsules for
experiments with a specific surface of 4 m2L–1 and 20 m2L–1, it
is found that the total coating mass per capsule is roughly the
same, which is possibly due to counteracting effects, i.e., the
capsule-TiO2 agglomeration increases but the amount of TiO2
per capsule decreases with the number of
capsules.
tions but after a certain pressure limit is achieved, the capsule
wall is broken and the tetradecane is compressed, leading to a
higher intrusion of mercury. Additional measurements are
required to clarify this point.
From a comparison of the extracted (Fig. 5, top) and calcined (Fig. 5, bottom) samples it can be seen that they show a
very similar behavior while the filling of the interstices between
the extracted capsules occurs at slightly smaller diameters.
Moreover, at diameters lower than 10 lm, the introduced volume for the extracted and calcined capsules increases up to a
diameter of ca. 1 lm and then stays constant without showing
an intermediate plateau. One possible explanation is that the
capsules are destroyed or form small cracks during the extrac-
1.2
Mercury Porosimetry
www.cet-journal.com
norm. spec. vol. [-]
Typical results of the mercury porosimetry
measurements for capsules after the synthesis and extraction (top) and calcination
(bottom) are given in Fig. 5. It can be seen
that for the capsules (Fig. 5, top) at low
pressure, i.e., for large pore diameters, the
normalized specific mercury volume increases with decreasing pore sizes. The interstices between the capsules are filled in
this range between 100–10 lm. Below this
size, when all the external voids are filled,
the imbibed amount of mercury stays constant. A further increase of the pressure
(filling of smaller pores) at ca. 0.015 lm
leads to a sharp rise of the volume of the
intruded mercury. It can be assumed that
pores in this size range can only be part of
the capsule wall, which would be supported by the fact that the tetradecane in the
calcinations step can evaporate without destroying the capsules. Moreover, on examination of the raw data for the measurement, it is found that the specific volume
intrusion in the corresponding range is
183.39 mm3g–1, which is 11.59 mm3 with
respect to a sample mass of 0.0632 g. The
capsule mass corresponds to a sample
volume of 78.65 mm3, which is considerably larger than the imbibed mercury.
However, from the density of the capsules
(0.8035 g cm–3), a weight percentage of the
polymer shell can be determined to be ca.
10 % of the capsules, which corresponds to
a mass of 6.32 · 10–3 g and a volume of
4.05 mm–3. Thus, the volume of the wall is
too small to contain a pore volume of
11.59 mm–3. Another possible explanation
is that the capsules in the pressure range
corresponding to pores larger than
0.15 lm are stable under isostatic condi-
1
0.8
0.6
0.4
0.2
capsules
2
-1
extrac. caps. 20 m L
extrac. caps. 4 m2 L-1
0
0.001
0.01
0.1
1
10
100
1
10
100
diameter [µm]
1.2
1
norm. spec. vol. [-]
3.3
0.8
0.6
0.4
0.2
2
-1
calc. caps. 20 m L
calc. caps. 4 m2 L-1
0
0.001
0.01
0.1
diameter [µm]
Figure 5. Hg-porosimetry of the capsules after synthesis and extraction (top) and after
calcination (bottom).
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
Hollow spheres
tion and calcinations or that the capsule wall cracks during the
measurement so that the mercury directly fills the inner spaces
of the particles. This assumption is supported by SEM investigations (see Fig. 6), where it can be seen that the walls of the
capsules are quite thin, especially for a specific surface of
20 m2L–1, and that a number of the capsules are partially broken.
Moreover, it is worth noting that in the cases of the extracted and calcined capsules, it seems that there is no increase
of the imbibed mercury below 0.5 lm. Since the capsules do
not contain any tetradecane, they cannot be compressed leading to an increase of the imbibed mercury. From a comparison
of the coating using a surface of 4 m2L–1 and 20 m2L–1, it can
be seen that in the case of a specific surface of 4 m2L–1, the
curve is shifted to smaller diameters, i.e., to higher pressure.
This again shows that thicker coatings are obtained for smaller
capsule surfaces.
SEM pictures of coated particles are shown in Fig. 6. Coated
capsules of a specific surface of 4 m2L–1 are shown in the two
top pictures. These show a dense uniform layer while larger
particles are incorporated in the coating. On the SEM picture
of the calcined capsules (top right), it can be seen that the inorganic layer is thick enough to form stable inorganic hollow
spheres. The coated capsules of a specific surface of 20 m2L–1
on the other hand show a coating layer that is smoother with a
lower amount of incorporated larger particles. Moreover, the
coating is considerably thinner leading to a higher amount of
broken capsules after the calcination.
4
Conclusions
Synthesized inorganic and organic-inorganic hybrid hollow
spheres were characterized by thermal analysis, infrared spec-
2035
troscopy and mercury porosimetry. Thermal analysis of the
capsules shows two endothermic effects, i.e., the loss of tetradecane between 100–250 °C and a small weight loss at ca.
400 °C which can, after comparison to the thermal analysis of
the pure resin, be assigned to the decomposition of the capsule
wall. In thermal analysis coupled with infrared spectroscopy, it
can be seen that the tetradecane evaporates at temperatures
below 100 °C while the main release takes place between 100–
250 °C. In addition, it is found that only carbon dioxide, small
amounts of carbon monoxide and water are formed in the
decomposition of the resin.
In the coated capsules, only the weight loss of the tetradecane and not the decomposition of the resin (due to the
small amount present) can be observed. In the same way, the
release of sulfur oxide from the precipitated titanium dioxide,
seen in the thermal analysis of pure TiO2, is not detectable in
the analysis of the coated capsules.
In the investigation of the percentage of TiO2 precipitated
on the capsules, it is observed that at a specific surface of
4 m2L–1, the percentage of precipitated TiO2 is considerably
higher in dilute sulfuric acid (24.9 wt-%) than in water
(14.0 wt %). This can be explained by the fact that in water at
the beginning of the experiment, there is a rise in pH leading
to a high nucleation rate and initial number density of TiO2particles. By taking into account that the precipitation is only
based on the agglomeration process between capsules and
TiO2-particles, the agglomeration of TiO2-particles with each
other counteracts the deposition of TiO2-particles on the capsules. Thus, in water there is initially an increased probability
for the agglomeration of TiO2-particles due to the higher
number of particles leading to a reduced coating of the capsules. The same effect can be seen for a specific surface of
20 m2L–1 (H2SO4: 15.3 wt-%; H2O: 13.2 wt-%) while the effect
is clearly lower since the mean distance between capsules and
TiO2-particles decreases while the mean distance
between the TiO2-particles stays roughly the same.
Another observation, which was expected, was that
the percentage of TiO2 on the capsules decreases
with an increase in specific capsule surface.
Samples were characterized by mercury porosimetry for determination of the porosity of the
capsules. It can be seen for capsules between 100–
10 lm that the interstices between the capsules are
filled with mercury while in the range of 10 to ca.
0.01 lm, the normalized specific volume stays constant showing no pores in this range. The observation that the imbibed volume increases again
afterwards is not yet completely understood but
can be possibly explained by the breakage of the
capsules and compression of the tetradecane. In
the case of the hollow spheres, it can be seen that
the mercury intrusion into the interstices between
the spheres cannot be separated from the filling of
the cavities in the spheres.
The authors have declared no conflict of interest.
Figure 6. SEM-pictures of coated and extracted (top and bottom left) and calcined (top and bottom right) capsules.
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cet-journal.com
2036
M. Sgraja et al.
References
[1] J. A. Hagarman, in Hollow and Solid Spheres and Microspheres: Science and Technology Associated With Their Fabrication and Application (Eds: D. L. Wilcox et al.), Materials
Research Society, Pittsburgh 1994.
[2] S. C. Zhang, G. L. Messing, S. Y. Lee, R. J. Santoro, in Hollow
and Solid Spheres and Microspheres: Science and Technology
Associated With Their Fabrication and Application (Eds: D. L.
Wilcox et al.), Materials Research Society, Pittsburgh 1994.
[3] J. Bertling, J. Blömer, R. Kümmel, Chem. Eng. Technol. 2004,
27 (8), 829.
[4] J. G. Liu, D. L. Wilcox Sr., J. Mater. Res. 1995, 10 (1), 84.
[5] B. Peng et al., J. Colloid Interface Sci. 2008, 321, 67.
[6] F. Caruso, Chem. Eur. J. 2000, 6 (3), 413.
[7] A. Syoufian, Y. Inoue, M. Yada, K. Nakashima, Mater. Lett.
2007, 61, 1572.
www.cet-journal.com
[8] J. Rong, J. Ma, Z. Yang, Macromol. Rapid Commun. 2004, 25,
1786.
[9] H. Strohm, P. Löbmann, J. Mater. Chem. 2004, 17, 2667.
[10] Y. Aoi, H. Kambayashi, E. Kamijo, J. Mater. Res. 2003, 18
(12), 2832.
[11] D. Wang, C. Song, Y. Lin, Z. Hu, Mat. Letters 2006, 60, 77.
[12] M. Sgraja, J. Blömer, J. Bertling, P. J. Jansens, J. Appl. Polym.
Sci. 2008, 110 (4), 2366.
[13] Handbook of Chemistry and Physics (Eds: D. R. Lide, H. P. R.
Frederikse), 78th ed., CRC Press, New York 1997.
[14] M. Leskelä, P. Eskelinen, M. Ritala, Thermochim. Acta 1993,
214, 19.
[15] W. F. Sullivan, S. S. Cole, J. Am. Ceram. Soc. 1959, 42 (3),
127.
[16] O. Saur et al., J. Catal. 1986, 99, 104.
© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eng. Technol. 2010, 33, No. 12, 2029–2036