Dynamics of Hollow and Solid Alumina Particle Formation in Spray Flames

Article abstract of DOI:10.1111/j.1551-2916.2004.00523.x

Thermophoretic sampling (TS) of the aerosol was conducted to manifest the formation of hollow and solid alumina particles in spray flames. The collected particles were investigated by transmission electron microscopy. Hollow particles with a thin shell (e.g., 10 nm) were formed from the aluminum nitrate precursor emulsion at less than 4-cm flame height. Hollow particles maintained their shapes in the flame using air as dispersion/oxidant gas, whereas hollow-to-solid restructuring of the particles took place in the flame using oxygen. With oxygen, nanoparticles were formed in the gas phase from the aluminum butoxide/2-propanol precursor solution only, whereas gas-phase reaction was hindered, forming large par ticles from the aluminum nitrate/2-propanol precursor solu tion.

Full text of DOI:10.1111/j.1551-2916.2004.00523.x

J. Am. Ceram. Soc., 87 [3] 523–25 (2004)
journal
Dynamics of Hollow and Solid Alumina Particle Formation in
Spray Flames
Takao Tani and Kazumasa Takatori†
Toyota Central Research and Development Laboratories, Inc., Nagakute, Aichi 480-1192, Japan
Sotiris E. Pratsinis
Particle Technology Laboratory, Department of Mechanical and Process Engineering,
Swiss Federal Institute of Technology (ETH) Zurich, CH-8092 Zurich, Switzerland
Thermophoretic sampling (TS) of the aerosol was conducted to
manifest the formation of hollow and solid alumina particles in
spray flames. The collected particles were investigated by
transmission electron microscopy. Hollow particles with a thin
shell (e.g., 10 nm) were formed from the aluminum nitrate
precursor emulsion at less than 4-cm flame height. Hollow
particles maintained their shapes in the flame using air as
dispersion/oxidant gas, whereas hollow-to-solid restructuring
of the particles took place in the flame using oxygen. With
oxygen, nanoparticles were formed in the gas phase from the
aluminum butoxide/2-propanol precursor solution only,
whereas gas-phase reaction was hindered, forming large par-
ticles from the aluminum nitrate/2-propanol precursor solu-
tion.
(AB)-derived alumina particles by the solution-fed FSP using
2-propanol as solvent.
II. Experimental Procedure
The AN nonahydrate (Wako, S grade) was dissolved in deion-
ized water, which was mixed with kerosene (Wako) and a
surfactant (hexa(2-hydroxy-1,3-propylene glycol) diricinoleate
(Taiyo Kagaku, Sunsoft 818H) at a volume ratio of 65 (AN
aqueous solution)/33 (kerosene)/2 (surfactant). The mixture was
stirred at 10 000 rpm for 10 min, forming the w/o emulsion of 1
mol/L precursor concentration, where aqueous microspheres of
1–2-␮m size were dispersed in the oil phase. The AN was
dissolved also in a mixture of 90 vol% 2-propanol (Wako, S grade)
and 10 vol% methanol (Wako, S grade) to obtain a 1 mol/L
solution. A 75 wt% AB/butanol solution (Gelest) was diluted with
2-propanol to 1 mol/L.
The detailed configuration of the FSP reactor was reported
elsewhere.10 The total methane and oxygen flow rates of the
supporting flames were 3 L/min each, while 12 mL/min precursor
solution was dispersed by air or oxygen of 11 L/min. Air of 45
L/min or oxygen of 15 L/min was supplied to the main flame
through a porous metal plate surrounding the nozzle for excess
oxidant when using air or oxygen, respectively, as dispersion gas.
Air was not used as dispersion/oxidant gas when dispersing the
2-propanol solutions because of an insufficient amount of oxidant.
A schematic of the equipment for TS is shown in Fig. 1. A copper
(Cu) mesh (3-mm diameter) covered with carbon film was set on
the edge of the sampling rod, which was driven by air of 2 atm at
1 m/s average velocity and held below the nozzle for 50 ms to
collect particles in the spray flame. The sampling time was
controlled by a timer setting. The TS was conducted at positions of
4, 6, and 8 cm below the nozzle along the central flame axis. It was
difficult to obtain particles at only 2 cm below the nozzle because
of melting of the Cu mesh. Even at sampling positions further
away, the carbon film was burned out, and therefore particles
attached with the Cu mesh were observed for some cases. The
particle morphology was observed by transmission electron mi-
croscopy (TEM: Nihon Denshi, JEM2000EX, 200 kV). The
particles are labeled by precursor (AN ϭ N or AB ϭ B)–process
(emulsion-fed FSP ϭ E or solution-fed FSP ϭ F)–dispersion/
oxidant gas (air ϭ A or oxygen ϭ O), so the nitrate-derived
alumina particles made by the emulsion-fed FSP using air as
dispersion/oxidant gas is labeled N-E-A.
I. Introduction
N
EMULSION-FED flame spray pyrolysis (FSP), the so-called
A
emulsion combustion method,1 is an elegant process to make
oxide particles by spraying and combusting a water-in-oil (w/o)
emulsion. It differs from the standard organic solution-fed FSP,2–4
because a high-content aqueous phase (e.g., 65 vol%) in the
precursor emulsion decreases the flame temperature, favoring
precipitation of the precursor in the liquid phase5 as in spray
pyrolysis (SP)6 rather than precursor evaporation and oxidation as
in the gas-phase route for particle formation.4
The emulsion-fed FSP of aluminum nitrate (AN) or aluminum
chloride using air as dispersion/oxidant gas made unique hollow
alumina particles which can be attractive for insulating and
lightweight filler materials as well as catalyst carriers because of
their small size (submicrometer), very thin shell (ϳ10 nm), and
high specific surface area (ϳ50 m2/g).5 Enhanced surface precip-
itation by combustion of the oil phase is considered to form the
hollow particles in both cases using the nitrate and chloride as
precursor. On the other hand, oxygen as dispersion/oxidant gas
increased the flame temperature, enhancing restructuring of the
hollow particles in the hot zone,7 resulting in solid particles from
the same AN precursor emulsion.5 In this study, thermophoretic
sampling (TS)8,9 of the aerosol in the spray flame was conducted
to demonstrate the evolution of hollow and solid alumina particle
formation in the emulsion spray flames. In addition, the results
were compared with those of the AN- or aluminum butoxide
A. Krell—contributing editor
III. Results and Discussion
Figure 2 shows the evolution of the alumina particle growth in
the emulsion spray flames (emulsion-fed FSP) using (a) air
Manuscript No. 10530. Received September 15, 2003; approved October 28, 2003.
†Author to whom correspondence should be addressed.
523

524
Communications of the American Ceramic Society
Vol. 87, No. 3
solid particles increased at 6 cm, and the powder consisted
primarily of solid particles at 8 cm, suggesting that the hollow
particles collapsed to solid ones7 in the spray flame for N-E-O as
presumed in the previous study.5
Figure 3 shows evolution of the alumina particle growth in the
2-propanol spray flames (solution-fed FSP) using (a) AB (B-F-O)
and (b) AN (N-F-O) as precursor. For B-F-O, large particles (e.g.,
200-nm diameter) and aggregates of nanoparticles were collected
at 4 cm, suggesting incomplete evaporation of the precursor
droplets as seen also in TS at 4 cm for zinc acetate-derived ZnO
and hexamethyldisiloxane-derived SiO2 particle synthesis.11 Only
aggregates of nanoparticles were observed at 6 and 8 cm, suggest-
ing complete evaporation of precursor droplets and particle for-
mation in the gas phase. The nanoparticles at 8 cm were about
10–30-nm diameter from the large-magnification TEM image. In
contrast, spherical particles of several hundred nanometers in
diameter were observed at 4 cm, and the particle morphology did
not change so much at 6 and 8 cm for N-F-O, suggesting that Al
species were not evaporated in the spray flame, forming large
particles corresponding to the dispersed droplets and/or their
fragments. In contrast to B-F-O-made alumina, endothermic oxi-
dation of AN may hinder its evaporation and thus gas-phase N-F-O
particle formation in the 2-propanol spray flame, as seen with
nitrate-derived bismuth oxide particles made by the solution-fed
FSP.12 Although some N-F-O particles seemed hollow with a thick
shell (e.g., 100 nm) judging from the contrast in the TEM images,
the unique thin-shell structures observed during N-E-A or N-E-O
particle growth were not seen for the N-F-O alumina particles.
These results suggest that the w/o emulsion phase of the precursor
solution droplets can result in thin-shell hollow particles by surface
precipitation or foaming at each of the aqueous microspheres of
the precursor emulsion.
Fig. 1. Schematic of the equipment for thermophoretic sampling of
aerosol in the spray flame.
(N-E-A) and (b) oxygen (N-E-O) as dispersion/oxidant gas. For
N-E-A, hollow particles with a thin shell (e.g., 10 nm) were
collected at 4 cm below the nozzle, indicating that formation of
hollow particles started at less than 4 cm. Only aggregates of
hollow particles were observed at 6 cm, while particle morphology
did not change further at 8 cm, although the particles had broad
size distributions. It was difficult to find explicitly the process
(e.g., surface precipitation or foaming) for hollow particle forma-
tion by the emulsion-fed FSP. On the other hand, aggregates of
hollow particles were seen at 4 cm where submicrometer-sized
solid and hollow particles coexisted for N-E-O. The fraction of the
Fig. 2. Evolution of the alumina particle growth from the aluminum
nitrate precursor emulsion along the axis of spray flames. Using air ((a)
N-E-A) as dispersion/oxidant gas resulted in only hollow particles with a
thin shell, whereas hollow-to-solid restructuring of the particles took place
in the flame using oxygen ((b) N-E-O) as dispersion/oxidant gas.
Fig. 3. Evolution of the alumina particle growth from the aluminum
butoxide ((a) B-F-O) or nitrate ((b) N-F-O)/2-propanol precursor solution
in the spray flames. Nanoparticles were formed in the gas phase for B-F-O,
whereas gas-phase reaction was hindered, forming large particles for
N-F-O.

March 2004
Communications of the American Ceramic Society
525
2M. Sokolowski, A. Sokolowska, A. Michalski, and B. Gokieli, “The ‘In-
Flame-Reaction’ Method for Al2O3 Aerosol Formation,” J. Aerosol Sci., 8,
219–30 (1977).
IV. Summary
Hollow particles with a thin shell (e.g., 10 nm) were formed at
less than 4 cm in flame height for N-E-A and N-E-O. It was
difficult to find explicitly the process for hollow particle formation
by the emulsion-fed FSP. The particle morphology did not change
so much in the flame for N-E-A, whereas hollow-to-solid restruc-
turing of the particles took place in the flame for N-E-O, resulting
in primarily solid particles at 8 cm. Nanoparticle formation in the
gas phase was observed in the flame for B-F-O. While, spherical
particles of several hundred nanometers in diameter were formed
in the flame for N-F-O, suggesting that endothermic oxidation of
AN may hinder its evaporation and thus particle formation in the
gas phase. The unique thin-shell structures observed in the
evolution of the N-E-A or N-E-O particle growth were not seen for
N-F-O, suggesting that the w/o emulsion phase of the precursor
solution droplets can result in unique hollow particles.
3R. M. Laine, T. Hinklin, G. Williams, and S. C. Rand, “Low-Cost Nanopowders
for Phosphor and Laser Applications by Flame Spray Pyrolysis”; pp. 500–10 in
Metastable, Mechanically Alloyed and Nanocrystalline Materials, Parts 1 and 2.
Trans Tech Publications Ltd., Zurich–Uetikon, Switzerland, 2000.
4L. Ma¨dler, H. K. Kammler, R. Mueller, and S. E. Pratsinis, “Controlled Synthesis
of Nanostructured Particles by Flame Spray Pyrolysis,” J. Aerosol Sci., 33 [2] 369–89
(2002).
5T. Tani, N. Watanabe, K. Takatori, and S. E. Pratsinis, “Morphology of Oxide
Particles Made by the Emulsion Combustion Method,” J. Am. Ceram. Soc., 86 [6]
898–904 (2003).
6G. L. Messing, S. C. Zhang, and G. V. Jayanthi, “Ceramic Powder Synthesis by
Spray Pyrolysis,” J. Am. Ceram. Soc., 76 [11] 2707–26 (1993).
7J. Ortega, T. T. Kodas, S. Chadda, D. M. Smith, M. Ciftcioglu, and J. E. Brennan,
“Formation of Dense Ba0.86Ca0.14TiO3 Particles by Aerosol Decomposition,” Chem.
Mater., 3 [4] 746–51 (1991).
8A. P. George, R. D. Murley, and E. R. Place, “Formation of TiO2 Aerosol from
the Combustion Supported Reaction of TiCl4 and O2,” Faraday Symp. Chem. Soc., 7,
63–71 (1973).
9O. I. Arabi-Katbi, S. E. Pratsinis, P. W. Morrison, and C. M. Megaridis,
“Monitoring the Flame Synthesis of TiO2 Particles by in-Situ FTIR Spectroscopy and
Thermophoretic Sampling,” Combust. Flame, 124 [4] 560–72 (2001).
10T. Tani, N. Watanabe, and K. Takatori, “Emulsion Combustion and Flame
Spray Synthesis of Zinc Oxide/Silica Particles,” J. Nanopart. Res., 5, 39–46
(2003).
Acknowledgments
The authors are grateful for experimental support from Mr. H. Morisaka (Toyota
CRDL). Design of the thermophoretic sampling equipment by Mr. Y. Osada (Toyota
CRDL) is acknowledged.
11T. Tani, K. Takatori, and S. E. Pratsinis, “Evolution of the Morphology of Zinc
Oxide/Silica Particles Made by Spray Combustion,” J. Am. Ceram. Soc., 87 [3]
365–70 (2004).
References
1K. Takatori, T. Tani, N. Watanabe, and N. Kamiya, “Preparation and Character-
ization of Nano-structured Ceramic Powders Synthesized by Emulsion Combustion
Method,” J. Nanopart. Res., 1, 197–204 (1999).
12L. Ma¨dler and S. E. Pratsinis, “Bismuth Oxide Nanoparticles by Flame Spray
Pyrolysis,” J. Am. Ceram. Soc., 85 [7] 1713–18 (2002).
Ⅺ