Effect of precursor and solvent on morphology of zirconia nanoparticles produced by combustion aerosol synthesis

Article abstract of DOI:10.1111/j.1151-2916.2003.tb00011.x

Zirconia nanoparticles were synthesized using a flame-based system involving spray droplet combustion of different precursor solutions. The characteristics of the feed were varied by varying the precursor compound, precursor concentration, and solvent type, and by using droplets of different mean sizes. When large droplets were used, agglomerated particles were formed when an organometallic precursor was used and large cenospheric particles were produced when an inorganic precursor was used. Reduction of the droplet size to a number-mean droplet diameter of 3.2 μm resulted in the production of solid spherical particles regardless of the precursor type. When an inertial impactor was used to eliminate droplets larger than 2.3 μm, the large particles in the final product were eliminated and uniformly sized solid zirconia particles having a smaller mean size were produced. The final particle size did not vary with the concentration of the precursor, indicating that multiple ceramic particles resulted from each precursor-containing droplet.

Full text of DOI:10.1111/j.1151-2916.2003.tb00011.x

J. Am. Ceram. Soc., 86 [2] 273–78 (2003)
journal
Effect of Precursor and Solvent on Morphology of Zirconia
Nanoparticles Produced by Combustion Aerosol Synthesis
Amit U. Limaye and Joseph J. Helble
Department of Chemical Engineering, University of Connecticut, Storrs, Connecticut 06269-3222
9
,21
Zirconia nanoparticles were synthesized using a flame-based
system involving spray droplet combustion of different precur-
sor solutions. The characteristics of the feed were varied by
varying the precursor compound, precursor concentration,
and solvent type, and by using droplets of different mean sizes.
When large droplets were used, agglomerated particles were
formed when an organometallic precursor was used and large
cenospheric particles were produced when an inorganic pre-
cursor was used. Reduction of the droplet size to a number-
mean droplet diameter of 3.2 ␮m resulted in the production of
solid spherical particles regardless of the precursor type.
When an inertial impactor was used to eliminate droplets
larger than 2.3 ␮m, the large particles in the final product
were eliminated and uniformly sized solid zirconia particles
having a smaller mean size were produced. The final particle
size did not vary with the concentration of the precursor,
indicating that multiple ceramic particles resulted from each
precursor-containing droplet.
and the concentration of the precursor in solution.
When large
droplets are used for precursor delivery in spray-pyrolysis systems,
cenospheric or shell-like hollow particles are produced because the
precursor and product diffusion rates in the droplet are generally
9
small compared with the solvent vaporization rate.
Precursor concentration in the solution also plays an important
role in defining final particle size and morphology, with low
concentrations resulting in small solid spherical particles. The
effects of precursor type and droplet size on the final particle
morphology have been investigated for spray-pyrolysis sys-
1
7,18,20,21
tems.
Similar data on high-temperature flame-based
droplet-combustion systems are, however, lacking.
In this study, the effects of different precursors, solvents, and
precursor solution droplet sizes on the final size and shape of the
nanoparticles produced at the higher heating rate condition of
flame synthesis are presented. In spray pyrolysis, each precursor
solution droplet produces one ceramic particle, thereby imposing a
requirement of low precursor concentrations for the formation of
solid particles. In contrast, flame-based droplet combustion results
in multiple particles per droplet enabling operation at higher feed
7
,8
I. Introduction
concentrations.
While the mechanism for the formation of
multiple ceramic particles per combustible droplet in droplet-
combustion systems is not well understood, the underlying pro-
cesses of particle formation are expected to be the same as in spray
pyrolysis. The competition between droplet vaporization and
intradroplet precursor diffusion has been shown to greatly affect
LAME-BASED methods for synthesis of nanoparticles are desir-
F
able because the processes are continuous, the product is free
of contamination, and postproduction processing is typically un-
1
–3
necessary.
While particles produced by flame-based methods
often have very small sizes (5–10 nm), the particles are usually
9
the final particle characteristics in spray-pyrolysis systems. The
3
agglomerated or “fractal-like.” Since branched structures are
3
4
typical heating rate in spray-pyrolysis systems of 10 –10 K/s is,
undesirable in many applications, attempts have been made to
control agglomeration in these particles but with moderate suc-
4
however, smaller than that for droplet-combustion systems (10 –
5
1
0 K/s). This suggests that in droplet-combustion systems, droplet
4
–6
cess.
However, it has been shown that unagglomerated spher-
vaporization dominates over the diffusion within the droplet,
resulting in multiple particles from one precursor droplet.
ical particles can also be produced in flame-based systems using
7
,8
droplet combustion methods. It has further been shown that the
same process can produce particles of either agglomerated or
7
spherical morphology by varying the flame temperature.
II. Experimental Procedure
Temperature, however, is not the only parameter affecting the
morphology of particles produced in flame-based systems. In
general, in high-temperature ceramic particle production pro-
cesses, the type of precursor plays an important role in determining
An experimental system designed to provide a controllable
temperature profile was used in the synthesis of ceramic nanopar-
ticles. This system consisted of a flat-flame porous-plug burner
supporting a premixed methane/oxygen/nitrogen flame, an atom-
izer for precursor solution delivery, and an insulated quartz reactor
for particle formation. Ceramic precursors were introduced into the
flame by spraying the precursor solution through a feed tube
located centrally in the burner. Details of the system have been
9
final particle characteristics such as morphology. For example,
gaseous, liquid, and solid precursors have all been used in
ceramic-oxide particle production using either flames or furnac-
1
0–19
es.
morphology is fractal-like due to the agglomeration of particles
When these precursors vaporize, the final ceramic particle
7
1
3,14
reported elsewhere. In this study two solvents were used: 99.8%
formed by condensation of the vapors.
When the precursors do
anhydrous butanol (Sigma Chemicals, Inc., St. Louis, MO) and
deionized ultrafiltered (DIUF) water (Fisher Scientific, Inc., Fair
Lawn, NJ). Two zirconium salts, zirconium(IV) n-butoxide (re-
ferred to here as ZrBu; Strem Chemicals, Inc., Newburyport, MA)
and zirconium(IV) dinitrate oxide hydrate (referred to here as
ZrNt; Strem Chemicals, Inc., Newburyport, MA) were used as
precursor solutes in the experiments. Zirconium(IV) n-butoxide is
water sensitive and undergoes rapid hydrolysis followed by con-
densation. Synthesis experiments were conducted with solutions of
zirconium(IV) dinitrate oxide hydrate in either water or butanol
and zirconium(IV) n-butoxide in butanol. While zirconium(IV)
dinitrate oxide hydrate is entirely soluble in water, 10% by volume
of pure ethanol (Aaper Alcohol Co. Shelbyville, KY) was used to
not vaporize but remain in the liquid state, however, particles form
in the liquid phase and are typically unagglomerated spheres.
The size of the spherical unagglomerated particles formed in
high-temperature systems is governed by the precursor droplet size
7
,17
T. M. Besmann—contributing editor
Manuscript No. 186884. Received June 24, 2002; approved November 6, 2002.
Supported by the National Science Foundation under CAREER Grant No.
CTS-9733234.
2
73

2
74
Journal of the American Ceramic Society—Limaye and Helble
Vol. 86, No. 2
stabilize it in butanol because of its partial solubility in this
solvent. In experiments using zirconium(IV) n-butoxide, the pre-
cursor was added to butanol with moderate stirring under nitrogen
in a glove box to avoid reaction with water vapor present in the air.
A pressure-driven atomizer (Model 9302, TSI, Inc., St. Paul,
MN) operating in the pressure range of 34.5–345 kPa was used to
produce the spray. In selected experiments, the spray was modified
by the inclusion of either a 2.3 ␮m cut-point impactor or a
gravitational settling chamber, both located between the atomizer
and the burner. The droplet size distributions of the spray produced
by the atomizer and the spray modified using the settling chamber
were characterized at the relevant operating pressures with a phase
doppler particle analyzer (PDPA) (Aerometrics, Inc., Sunyvale,
CA). Both the impactor and the settling chamber served to
eliminate the large droplets, thus enabling control of the spray
droplet sizes. An operating pressure of 41.5 kPa was used in
experiments involving unmodified feed spray, while the atomizer
was operated at 69 kPa in experiments involving the settling
chamber. The experiments involving the impactor required an
operating pressure of 172 kPa.
using larger droplet sizes was not, however, investigated. Larger
droplet sizes were therefore examined by using an atomizer that
produced a broad spread in the droplet sizes and consequently a
larger mean droplet size. The atomizer used in this latter study had
a number-mean droplet size of 7 ␮m with 17% of the droplets
being larger than 10 ␮m, whereas the atomizer used in the earlier
study produced a spray with a number-mean droplet size of 6 ␮m
with only 11% of the droplets being larger than 10 ␮m. The size
distribution of the droplets produced by the atomizer used in this
study is indicated by the curve labeled “unmodified” in Fig. 1.
When the spray was modified using the settling chamber, the
relatively large droplets were removed, resulting in a number-
mean droplet size of 3.2 ␮m with 2.5% of the droplets being larger
than 10 ␮m. This is seen in the curve labeled “settling chamber”
in Fig. 1. The droplet size distribution of the spray when the
impactor was used could not be measured with the PDPA due to
inadequate signal strength resulting from the reduced mass flow
rate. The curves in Fig. 1 have been normalized by the total
number count of droplets to provide a fractional distribution of
droplet sizes.
Solutions having three different concentrations of the precur-
sor—0.016M, 0.047M, and 0.141M—were prepared with each
solvent. The resulting solutions are indicated in Table I and are
referred to as solutions A through I. Each of the experiments A
through I was conducted at three conditions: unmodified spray,
spray modified by the settling chamber, and spray modified by the
impactor. This permitted the investigation of droplet size on the
final particle characteristics for all possible combinations of
precursors and solvents.
Introduction of the precursor droplets into the flame produced
ceramic nanoparticles that were analyzed for size and shape using
the SMPS and TEM. Results indicated that when the unmodified
spray was used, agglomerated or fractal-like ceramic particles
were the dominant ceramic particle morphology. These fractal-like
particles were eliminated, however, when the large droplets in the
spray were eliminated using the settling chamber. This is seen in
the TEM micrographs of selected particles produced in both cases
and shown in Figs. 2(a) and (b).
On introduction of the precursor solution into the flame,
solvent vaporization and particle formation occurred in the
reactor. The resulting nanoscale particles were collected down-
stream of the flame using a movable water-cooled, nitrogen-
quenched collection probe. A point-to-plane electrostatic pre-
cipitator (ESP) (Intox Products, Inc., Albuquerque, NM) was
used to separate the particles from the gas stream by deposition
onto a transmission electron microscope (TEM) grid. A scan-
ning mobility particle sizer (SMPS) (TSI, Inc., St. Paul, MN)
was also used to obtain on-line measurement of particle size
distributions in the range of 18–900 nm. Particles deposited on
the TEM grid were analyzed using a TEM (Model 420, Phillips,
Eindhoven, The Netherlands) to obtain information on particle
morphology, size, and crystallinity.
The particular morphology—spherical or fractal-like—that
dominates the ceramic particles results from the competition
between two distinctly different pathways of particle formation,
one through the vapor phase and the other through the liquid
7
,14
phase.
Two important parameters governing the dominant
pathway of particle formation are reactor temperature and precur-
sor volatility. Results shown in Fig. 2 indicate that droplet size also
contributes. Droplet size, however, is not an independent param-
eter but is tied to the system temperature profile. As described
previously, the precursor feed was introduced through a feed port
located in the center of the burner. Consequently, a core region
exists in the center of the flame where the combustion of the
solvent in the precursor solution contributes as a significant heat
source. This heat of combustion is expected to affect the particle
III. Results and Discussion
In all experiments in this study, a steady state temperature
profile was obtained in the reactor within 20–25 min of igniting
the burner gases. The premixed flame was run at an equivalence
ratio of one (␸ ϭ 1.0) based on the methane and oxygen content.
All experiments used a flame having a calculated adiabatic
temperature of 1710 K resulting from a premixed 1/2/13.2 meth-
ane/oxygen/nitrogen mixture. The precursor solution was then
sprayed through the central feed tube into the core region of the
flame using the atomizer described previously.
ZrBu in butanol has been used in an earlier study to produce
particles of both agglomerated and unagglomerated spherical
7
morphologies. In this earlier effort, it was found that decreasing
the mean droplet size of the precursor solution spray from 6 ␮m to
less than 2.3 ␮m through inertial impaction resulted in a corre-
sponding decrease in the mean zirconia particle size. The effect of
Table I. Composition of the Different Solutions Used for
Particle Synthesis in the Study
0.141M
ZrBu
0.047M
ZrBu
0.016M
ZrBu
0.141M
ZrNt
0.047M
ZrNt
0.016M
ZrNt
Solvent
Butanol
Water
†
A
†
B
†
C
†
D
G
E
H
F
I
Fig. 1. Size distributions of butanol droplets with and without the settling
chamber. The size distribution is normalized by the total number of
droplets counted in each case to provide a fractional size distribution.
Not applicable.

February 2003
Morphology of Zirconia Nanoparticles Produced by Combustion Aerosol Synthesis
275
Fig. 2. TEM images of (a) fractal-like and spherical particles formed when large unmodified droplets of ZrBu in butanol are used in particle synthesis (the
size bar is 240 nm) and (b) the absence of fractal-like agglomerates when the settling chamber is used (the size bar is 140 nm).
formation process in its early stages. In this study, the unmodified
spray was obtained at an atomizer operating pressure of 41.5 kPa
and corresponds to a precursor solution delivery rate of 8 mL/h.
Combustion of the solvent (butanol) leads to a local temperature of
morphology can have different electrical mobility, one cannot
reliably use the SMPS to obtain particle size distribution for
experiments in which both fractal-like and unagglomerated spher-
ical particles are produced. Consequently, data from only those
runs involving the production of unagglomerated spherical parti-
cles are reported in Table II. As can be seen from Table II, the
mean particle size was found to increase by a factor of 1.6 when
the precursor concentration was increased by an order of magni-
tude in the experiments involving the settling chamber. In addition,
the relative number of large particles in the final product also
decreased with decreasing concentration as shown by the fraction
of particles larger than 160 nm. For cases where the impactor was
used to eliminate the large droplets, the final mean particle size did
not change when the concentration of the precursor was changed
by nearly an order of magnitude.
1
290 K in the core region based on calculated adiabatic flame
temperature. The use of settling chamber reduced the number of
large droplets and thus the rate of solvent delivery into the reactor.
This thereby reduced the amount of heat released in the core
region. The calculated local temperature of the combustion prod-
ucts and the carrier gas in this case was 860 K. Since the volatility
of ZrBu is a strong function of temperature, this small variation in
the temperature changes the vaporization rate of the precursor
2
2
(
ZrBu). The vapor-phase pathway favored at high temperatures
therefore becomes dominant when the feed is not modified. This
trend of formation of fractal-like particles with the largest droplets
is observed at all concentrations of ZrBu considered (experiments
A, B, and C).
While the use of the settling chamber resulted in the elimination
of the fractal-like particles, the use of 2.3 ␮m cut-point impactor
resulted in the formation of even smaller unagglomerated spherical
particles having a number-mean size of 65 nm. A summary of the
particle sizes produced using different experimental conditions is
provided in Table II. The data reported in Table II were obtained
from the SMPS and thus depended on the electrical mobility of the
particles. Since particles having the same mass but different
The type of precursor also affected the final ceramic particle
morphology. This was demonstrated by the results of the experi-
ments using ZrNt in butanol. Observations indicated that when the
unmodified spray was used, a mixture of large cenospheric
particles and smaller solid ones was produced. Fragments of
cenospheres and heavily deformed shell-like structures were also
seen in the final material. Figure 3 shows TEM micrographs of (a)
solid, (b) hollow, (c) shell-like, and (d) deformed shell-like
particles obtained from experiment E using unmodified droplets.
The images in Fig. 3 are shown at a comparable magnification to
demonstrate the wide variation of overall feature size in these
particles. Figure 3 also demonstrates that the shell-like particles
were composed of small crystallites. Similar observations have
been made previously for magnesium oxide particles produced
Table II. Number-Mean Particle Size, and Fraction of
Zirconia Particles Larger Than 160 nm for Particles
Produced under Different Experimental Conditions
2
3
from magnesium nitrate in spray pyrolysis systems. The pres-
ence of cenospheric shell-like particles is consistent with the
hypothesis that the rate of droplet solvent vaporization is much
faster than that of intraparticle solute diffusion in the ZrNt/butanol
system. When the droplet is large and the surface recedes because
of solvent vaporization, the local concentration of the precursor
increases. Surface precipitation occurs when the saturation con-
centration is exceeded. The solvent inside the precipitated shell
continues to vaporize and escapes through the pores in the shell,
resulting in an intact porous particle. If, however, the shell is
impermeable to the vapors, the shell fractures because of the
buildup of internal pressure, resulting in shell-like fragmented
particles. It has also been suggested that since ZrNt melts before
Settling chamber
Inertial impactor
Run
Mean (nm)
% Ͼ160 nm
Mean (nm)
% Ͼ160 nm
A
B
C
D
E
F
G
H
I
140
90
26
11
8
14
10.5
5
19
15
9.5
69
63
63
86
77
56
75
61
49
4.3
0.2
0.6
5
1.5
0.2
3.5
0.4
0.1
86
103
94
71
114
100
82

2
76
Journal of the American Ceramic Society—Limaye and Helble
Vol. 86, No. 2
Fig. 3. TEM images of particle morphologies obtained when large unmodified droplets of ZrNt in butanol are used in particle synthesis: (a) solid (the size
bar is 275 nm), (b) hollow (the size bar is 275 nm), (c) shell-like (the size bar is 275 nm), and (d) deformed shell-like particles (the size bar is 240 nm).
transformation into zirconia, deformed structures are also observed
in the final product. When small droplets are used, however,
used. Elimination of cenospheric particles as the precursor droplet
size decreased was also seen at these high precursor concentra-
tions. The effect of precursor concentration on the final particle
sizes when the feed was modified using the impactor is indicated
in Fig. 4. Comparison of the results in Table II and Fig. 4 indicates
that the mean particle size increased by a factor of 1.4–1.5 from
case F to D. If one particle resulted from one droplet, the mean
particle size in experiment D would be expected to be 2.1 times the
mean particle size in experiment F since the concentration of the
precursor in the experiment D was 9 times that in experiment F.
The factor of 1.5 increase indicates multiple particle formation per
precursor solution droplet. The data in Table II indicate that in
addition to a decrease in the mean particle size, the contribution of
large particles in the final product also decreased with a reduction
in concentration, as indicated by the reduced fraction of particles
that were larger than 160 nm.
9
diffusion distances within the droplet are smaller, thereby allowing
an increase in local solute concentration in the interior of the
droplet. Precipitation consequently occurs at multiple locations
within the droplet to form small solid particles. This is consistent
with the observation here of solid particles being formed when the
settling chamber was used to eliminate the large droplets. On
further reduction of the mean spray droplet size with the impactor,
the final particles were found to be spherical, solid, and uniformly
sized.
Given these observations, it is expected that higher precursor
concentrations in large droplets would result in increased shell
formation. TEM observation of the particles formed in run D
revealed the presence of large shell-like and deformed shell-like
particles when the large droplets from the unmodified feed were

February 2003
Morphology of Zirconia Nanoparticles Produced by Combustion Aerosol Synthesis
277
Fig. 4. Size distributions measured by SMPS of spherical zirconia
nanoparticles produced using the 2.3 ␮m impactor with different concen-
trations of the precursor ZrNt in butanol. Each curve represents the average
of three experiments. Error bars represent standard deviation of the mean
value of the number distribution normalized by the total number count (NT)
at the indicated point.
Fig. 5. Size distributions measured by SMPS of spherical zirconia
nanoparticles produced using the 2.3 ␮m impactor with different concen-
trations of the precursor ZrNt in water. Each curve represents the average
of three experiments. Error bars represent standard deviation of the mean
value of the number distribution normalized by the total number count (NT)
at the indicated point.
It is clear that the precursor does play an important role in the
final particle morphology. ZrBu vaporizes on heating when large
droplets containing ZrBu are used as the feed. The vapor-phase
pathway is therefore dominant when ZrBu is the precursor. In
contrast, when ZrNt is used with large droplets, shell-like zirconia
particles are formed. It has been suggested in the spray pyrolysis
literature that when alkoxides are used in particle synthesis, they
form hydroxides, oxides, and intermediates due to hydrolysis and
significant effect of the solvent for a given precursor concentra-
tion. The trend of a reduction in the number of large particles
(diameter Ͼ 160 nm) in the final product with decreasing concen-
tration was also seen in this case.
IV. Sumary and Conclusions
1
8,24
Droplet combustion synthesis of zirconia nanoparticles has been
examined to determine the effect of parameters such as droplet
size, precursor concentration, precursor and solvent type on the
final ceramic nanoparticle morphology and size. When zirconi-
um(IV) n-butoxide in butanol was used, large droplets resulted in
the production of fractal-like particles. These particles were
eliminated on the removal of the large droplets from the feed. The
formation of fractal-like particles only when large droplets were
used was attributed to higher local temperatures in the particle
formation zone resulting from the higher mass flow rate of
combustible solvent. When zirconium(IV) dinitrate oxide hydrate
in butanol was used, however, the presence of large droplets
resulted in the formation of hollow and shell-like particles. This
was attributed to the relatively low volatility of the precursor and
to surface precipitation within the droplet due to the attainment of
local saturation conditions. Solid particles were obtained when the
large droplets were eliminated using a settling chamber.
The solvents used in this study did not affect the final particle
morphology and size. The observations made in this study on the
effect of precursor concentration and droplet size are consistent
with observations made in the spray pyrolysis literature. Though
final particle sizes were found to be smaller for lower precursor
concentrations, the decrease in the mean particle size did not
reflect the concentration of the precursor in the solutions indicating
the formation of more than one particle from each droplet. This
finding is contrary to the predictions of spray pyrolysis and is
explained by the higher heating rate in the droplet combustion
system compared with a typical spray pyrolysis system.
polymerization while in solution.
This has been termed vol-
ume gelation or polymerization. Consequently, small pockets of
precursors may form local networks that precipitate and crystallize
on solvent vaporization, resulting in the formation of solid parti-
cles rather than shell-like particles. This also explains the relative
uniformity of the particle sizes irrespective of the concentration for
solutions A–C as opposed to the variation observed with solutions
D–F when using the impactor as indicated in Table II.
While the above results clearly show a dependence of final
particle morphology on precursor type, concentration, and size of
the feed droplets, the volatility of the solvent is also expected to
play a role in the particle formation process. Experiments G
through I were conducted to assess the effect of solvent volatility.
Because butanol is less volatile than water, the rate of water
droplet vaporization is expected to lead to precipitation earlier in
the droplet lifetime. For example, a water droplet is calculated to
vaporize 4 times faster than a butanol droplet of the same size at
an ambient temperature of 800 K. This difference in volatility is
expected to translate into increased formation of shell-like and
hollow particles. TEM observations for runs G–I indicated that
when the unmodified spray was used to produce particles, shell-
like particles were observed irrespective of the precursor concen-
tration. A comparison of the size of the fragments and the
shell-like particles could not be done because of the spread in the
sizes of these particles. No conclusion regarding the relative size or
the number of shell-like particles could therefore be drawn from
the TEM data. TEM observations, however, indicated that when
the settling chamber was used, the shell-like particles were
eliminated. On further reduction of the droplet size using the
impaction chamber, the final particles were uniformly sized and
solid. The size distributions of the final particles obtained from
runs G–I when the impactor was used to modify the feedstream are
indicated in Fig. 5. Comparison of the curves in Fig. 5 and the data
for solutions G–I in Table II indicates that the relative difference
in the initial feed concentration was not reflected in the final mean
particle sizes, clearly demonstrating the formation of multiple
particles per droplet. Comparison of particle size data for solutions
D–F with G–I from Table II also indicates that there was no
Acknowledgment
We thank Dr. Donald Potter, Department of Metallurgy and Materials Engineer-
ing, University of Connecticut, Storrs, for stimulating discussions regarding electron
microscopy.
References
1
D. E. Rosner, “Combustion Synthesis and Materials Processing,” Chem. Eng.
Educ., [Fall] 228–35 (1997).

2
78
Journal of the American Ceramic Society—Limaye and Helble
Vol. 86, No. 2
2
14
G. D. Ulrich, “Flame Synthesis of Fine Particles,” Chem. Eng. News, [August 6]
J. J. Helble, “Combustion Aerosol Synthesis of Nanoscale Ceramic Powders,”
2
2–29 (1984).
J. Aerosol Sci., 29 [5–6] 721–36 (1998).
3
15
S. E. Pratsinis, “Flame Aerosol Synthesis of Ceramic Powders,” Prog. Energy
J. Karthikeyan, C. C. Berndt, J. Tikkanen, J. Y. Wang, A. H. King, and H.
Combust. Sci., 24 197–219 (1998).
S. Vemury and S. E. Pratsinis, “Charging and Coagulation During Flame
Herman, “Nanomaterial Powders and Deposits Prepared by Flame Spray Processing
4
of Liquid Precursors,” Nanostuct. Mater., 8 [1] 61–74 (1997).
1
6
Synthesis of Silica,” J. Aerosol Sci., 27, 951–66 (1996).
R. Baranwal, M. P. Villar, R. Garcia, and R. M. Laine, “Flame Spray Pyrolysis
5
of Precursors as a Route to Nanomullite Powder: Powder Characterization and
M. Katzer, A. P. Weber, and G. Kasper, “Effect of Electrical Fields on the Growth
Sintering Behavior,” J. Am. Ceram. Soc., 84 [5] 951–61 (2001).
of Titania Particles Formed in a Methane–Oxygen Diffusion Flame,” J. Aerosol Sci.,
1
7
I. W. Lenggoro, T. Hata, F. Iskander, M. M. Lunden, and K. Okuyama, “An
3
2, 1045–67 (2001).
6
Experimental and Modeling Investigation of Particle Production by Spray Pyrolysis
D. Lee and M. Choi, “Coalescence Enhanced Synthesis of Nanoparticles to
Using a Laminar Flow Aerosol Reactor,” J. Mater. Res., 15 [3] 733–43 (2000).
Control Size, Morphology and Crystalline Phase at High Concentrations,” J. Aerosol
Sci., 33, 1–16 (2002).
1
8
G. L. Messing and S. C. Zhang, “Ceramic Nanoparticle Synthesis by Spray
7
Pyrolysis”; pp. 185–90 in Euro-Ceramics II. Edited by G. Ziegler and H. Hausner.
A. U. Limaye and J. J. Helble, “Morphological Control of Zirconia Nanoparticles
Deutsche Keramische Gesellschaft, Cologne, Germany, 1991.
Through Combustion Aerosol Synthesis,” J. Am. Ceram. Soc., 85 [5] 1127–32 (2002).
19
8
I. S. Altman, “On Condensation Growth of Oxide Particles During Gas-Phase
A. Kilian and T. F. Morse, “A Novel Aerosol Combustion Process for the High
Combustion of Metals,” Combust. Sci. Technol., 160, 221–29 (2000).
Rate Formation of Nanoscale Oxide Particles,” Aerosol Sci. Technol., 34 227–35
20
G. L. Messing, S. C. Zhang, and G. V. Jayanti, “Ceramic Powder Synthesis by
(
2001).
Spray Pyrolysis,” J. Am. Ceram. Soc., 76 [11] 2707–26 (1993).
9
S. C. Zhang and G. L. Messing, “Synthesis of Solid, Spherical Zirconia Particles
by Spray Pyrolysis,” J. Am. Ceram. Soc., 73 [1] 61–67 (1990).
21
2
S. C. Zhang, G. Mulholland, and G. L. Messing, “Synthesis of ZrO Nanopar-
ticles by Spray Pyrolysis,” J. Mater. Synth. Process., 4 [4] 227–33 (1996).
1
0
S. E. Pratsinis, Z. Wenhua, and S. Vemury, “Role of Gas Mixing in the Flame
Synthesis of Titania Powders,” Powder Technol., 86 [1] 87–93 (1996).
22
D. C. Bradley, R. C. Mehrotra, J. D. Swanwick, and W. Wardlaw, “Structural
Chemistry of the Alkoxides: Part IV, Normal Alkoxides of Silicon, Titanium and
1
1
O. I. Arabi-Katbi, S. E. Pratsinis, P. W. Morrison Jr., and C. M. Megaridis,
Monitoring the Flame Synthesis of TiO Particles by In Situ FTIR Spectroscopy and
Thermophoretic Sampling,” Combust. Flame., 124, 560–72 (2001).
Zirconium,” J. Chem. Soc., 2025–30 (1953).
“
2
23T. J. Gardner and G. L Messing, “Magnesium Salt Decomposition and Morpho-
logical Development During Evaporative Decomposition of Solutions,” Thermochim.
Acta, 78, 17–27 (1984).
1
2
H. D. Jang, “Generation of Silica Nanoparticles from Tetraethyl Orthosilicate
TEOS) Vapor in a Diffusion Flame,” Aerosol Sci. Technol., 30, 477–88 (1999).
2
4
(
2 3
H. Ishizawa, O. Sakurai, N. Mizutani, and M. Kato, “Homogeneous Y O -
1
3
B. Xia, L. Duan, and Y. Xie, “ZrO
2
Nanopowders Prepared by Low-Temperature
Stabilized ZrO2 Powder by Spray Pyrolysis Method,” Am. Ceram. Soc. Bull., 65 [10]
1399–404 (1986).
Vapor-Phase Hydrolysis,” J. Am. Ceram. Soc., 83 [5] 1077–80 (2000).
Ⅺ