Nanocrystallization and phase transformation in monodispersed ultrafine zirconia particles from various homogeneous precipitation methods

Article abstract of DOI:10.1111/j.1151-2916.1999.tb02085.x

Monodispersed ultrafine (nano- to micrometer) zirconia precursor powders were synthesized by three different physicochemical methods: (I) forced hydrolysis, (II) homogeneous precipitation in inorganic salt solutions, and (III) hydrolysis/condensation of alkoxide. The forced hydrolysis method produced monoclinic nanocrystalline particles (cube shaped) of nanometer scale, which depended largely on the initial salt concentration. Methods II and III, both involving the use of alcohol as a solvent, exhibited a faster particle formation rate and generated amorphous ultrafine (submicrometer) monodispersed microspheres, indicating that the presence of alcohol may have stimulated particle nucleation due to its low dielectric property (and, thus, the low solubility of nucleus species in mixed water-alcohol solutions). Nucleation and growth of the particles in solutions are discussed based on the measurements obtained by small-angle X-ray scattering (SAXS) and dynamic light scattering (DLS). High-temperature X-ray diffraction (HTXRD) and TGA/DTA studies elucidated the differences in phase transformation for different types of powders. The most interesting finding was the nonconventional monoclinic nanocrystal nucleation and growth that occurred prior to transformation to the tetragonal phase (at 1200°C) during the heat treatment of the nanocrystalline powders produced by the forced hydrolysis.

Full text of DOI:10.1111/j.1151-2916.1999.tb02085.x

J. Am. Ceram. Soc., 82 [9] 2313–20 (1999)
ournal
J
Nanocrystallization and Phase Transformation in
Monodispersed Ultrafine Zirconia Particles from
Various Homogeneous Precipitation Methods
,
†
†
‡
,‡
Michael Z.-C. Hu,* Rodney D. Hunt, E. Andrew Payzant, and Camden R. Hubbard*
Chemical Technology Division and Metals and Ceramics Division, Oak Ridge National Laboratory,
Oak Ridge, Tennessee 37831
Monodispersed ultrafine (nano- to micrometer) zirconia
precursor powders were synthesized by three different
physicochemical methods: (I) forced hydrolysis, (II) homo-
geneous precipitation in inorganic salt solutions, and (III)
hydrolysis/condensation of alkoxide. The forced hydrolysis
method produced monoclinic nanocrystalline particles
morphology/shape, microstructure, composition purity, crystal-
linity, etc. These characteristics of the precursor powders sig-
nificantly affect the later-stage processing and sintering prop-
erties and eventually determine the microstructure and
performance of the final ceramic materials. For example,
spherical particle shape and narrow distribution of size help to
reduce serious microstructural defects in ceramics (due to the
particle packing heterogeneities) by enhancing the powder flow
(
cube shaped) of nanometer scale, which depended largely
on the initial salt concentration. Methods II and III, both
involving the use of alcohol as a solvent, exhibited a faster
particle formation rate and generated amorphous ultrafine
1
and packing efficiency in green bodies.
Of particular interest to this study was the synthesis and
characterization of monodispersed zirconia precursor particles,
which are rather important in the fabrication of various modern
(
submicrometer) monodispersed microspheres, indicating
that the presence of alcohol may have stimulated particle
nucleation due to its low dielectric property (and, thus, the
low solubility of nucleus species in mixed water–alcohol
solutions). Nucleation and growth of the particles in solu-
tions are discussed based on the measurements obtained by
small-angle X-ray scattering (SAXS) and dynamic light
scattering (DLS). High-temperature X-ray diffraction
2
zirconia-based or zirconia-containing ceramics. Increasing in-
terest in zirconia ceramics may be attributed to their unique set
of properties, which includes high refractoriness and corrosion
resistance, good mechanical strength, high fracture toughness
and hardness, ionic conduction, high melting point, low ther-
mal conductivity at high temperature, and thermal stability and
3
(HTXRD) and TGA/DTA studies elucidated the differences
resistance to thermal shock. Zirconia is an important compo-
in phase transformation for different types of powders. The
most interesting finding was the nonconventional mono-
clinic nanocrystal nucleation and growth that occurred
prior to transformation to the tetragonal phase (at 1200°C)
during the heat treatment of the nanocrystalline powders
produced by the forced hydrolysis.
nent in diverse structural ceramics and specialty ceramics with
potential applications such as the following: (1) advanced
structural/transformation-toughened ceramics (wear parts such
as bearings, shapes, and seals; engine and machine compo-
nents; cutting and abrasive tools; mill media; refractories), (2)
electronic components (PZT piezoelectric ceramics and zir-
conate dielectric ceramics), (3) sensors/probes (for oxygen
transport and detection) and solid electrolytes (for fabrication
of fuel cells), (4) catalytic materials or supports with unique
properties, (5) pigments (ceramic colors), and (6) thermal bar-
rier coatings or thermal shield.
I. Introduction
DVANCED ceramic materials are in high demand for use in
A
structural or functional applications. Monodispersed ultra-
As such, an intense interest has grown in preparing zirconia
powders with controlled particle size and crystal structure. Di-
verse physical and chemical synthesis routes have been re-
ported in the literature for ultrafine zirconia particle produc-
tion. Low-temperature aqueous chemical processing has been
the leading contender among the low-cost processes for ultra-
fine (i.e., submicrometer to nanosize) ceramic particles are im-
portant building blocks used in the fabrication of advanced
ceramic materials having improved properties (mechanical,
electrical, optical, magnetic, thermal, ionic conducting, and
catalytic). Production of ceramic powders is by no means a
new research subject; however, practical methods are still
needed for synthesizing high-quality ultrafine powders with
required characteristics in terms of their size, size uniformity,
4
fine ceramic powder synthesis. In particular, hydrolysis (and
condensation) under hydrothermal conditions and homoge-
neous precipitation are common approaches. Typical methods
using the hydrolysis approach include forced hydrolysis in
5
–7
aqueous solutions of inorganic salt
and hydrolysis/
8–12
condensation of metal alkoxides in alcohol solutions.
An
L. C. Klein—contributing editor
organic solvent such as alcohol is always involved in the alk-
oxide hydrolysis/condensation method. Recently, alcohol sol-
vent was also reported to play an important role in inducing
homogeneous precipitation of hydrous oxide particles in aque-
Manuscript No. 190007. Received July 22, 1998; approved March 1, 1999.
This work was supported by the U.S. Department of Energy, Basic Energy Sci-
ences, Division of Materials Sciences. The HTXRD experiments were conducted in
the High Temperature Materials Laboratory of Oak Ridge National Laboratory, spon-
sored by the Assistant Secretary for Energy Efficiency and Renewable Energy, Office
of Transportation Technologies, as part of the High Temperature Materials Labora-
tory User Program, Oak Ridge National Laboratory, managed by Lockheed Martin
Energy Research Corp. for the U.S. Department of Energy under Contract No. DE-
AC05-96OR22464.
5
ous solutions of inorganic salts.
Formation of hydrous-zirconia particles by the above-
mentioned methods has been studied extensively. However,
characterization of the ultrafine powder product, particularly
comparative studies of the monodispersed particles produced
from different particle synthesis processes, has not been re-
ported in previous studies. One of the important issues is to
understand the nanocrystal evolution/nucleation, growth, and
*
Member, American Ceramic Society.
†
Chemical Technology Division.
Metals and Ceramics Division.
‡
2313

2314
Journal of the American Ceramic Society—Hu et al.
Vol. 82, No. 9
crystalline-phase transformation during the heat treatment of
zirconia powders. The critical parameters that need to be con-
trolled are grain size and phase transformation of zirconia as a
second phase in order to ensure that the zirconia-containing
materials have the required properties. So far, most studies in
crystallization and phase transformation of zirconia have been
conducted on poor-quality particles or gel materials prepared
by simple chemical precipitation.
This study focuses on the characterization of monodispersed
powder products produced by the three particle-forming pro-
cesses. Comparison studies of particle-synthesis processes are
also provided. In situ high-temperature X-ray diffraction
description of the DLS spectrometer and the SAXS instrument
is provided by Hu et al.
5
,6
Several techniques were used for powder characterizations
and calcination. A scanning electron microscope (SEM) (Joel
JSM-T220A) and a Jeol 100CXII transmission electron micro-
scope (TEM) were used to image the size, morphology, and
microstructure of the powder particles (i.e., sol particles dried
at room temperature), while the powder’s surface area was
measured by the multipoint BET nitrogen adsorption technique
(Gemini, Micromiretics, Georgia). Powder samples were ana-
lyzed with a Harrop Model ST-736 thermogravimetric ana-
lyzer/differential thermal analyzer (TGA/DTA) for preliminary
phase-transition information. The total gas flow in each experi-
ment was 100 mL/min (20% O and 80% N ), which was
1
3
(
HTXRD) studies showed variation in the behavior of zirconia
crystallization and phase transformation for powders obtained
from various synthesis methods. Knowledge of the nanocrystal
formation characteristics and phase transformation can help in
determining the optimal thermal treatment process used in the
fabrication of advanced zirconia-based ceramics.
2
2
regulated by mass flow controllers from MKS Industries, Inc.
Approximately 0.167 g of powder sample was transferred into
an alumina sample holder, which was then placed into the
TGA/DTA. The sample was heated from ambient to 1200°C at
a rate of 200°C/h.
An in situ HTXRD instrument at the ORNL HTML user
facility was utilized for extensive studies of powder crystalli-
zation and phase transition. The HTXRD work was carried out
using a Scintag PAD X ␪–␪ diffractometer which was outfitted
with a constant-potential X-ray generator operating 45 kV/40
mA with a Copper target (␭CuK␣ ס 1.540598 Å), a Bueler
high-temperature XRD furnace with a beryllium window, a
thin Pt/Rh heating strip, and a liquid-nitrogen-cooled germa-
nium single-crystal solid-state detector. There are several sig-
nificant advantages of using a ␪–␪ goniometer: (1) the speci-
men is on a horizontal heating strip and is undisturbed by the
goniometer movement, (2) there is no need for the high-
temperature furnace to rotate, and (3) the same volume of
II. Experimental Procedure
High-purity zirconium salt (ZrOCl иxH O, 99.99%, FW
2
2
1
78.3), technical-grade zirconium tetrabutoxide in 80% butyl
alcohol, and 200 proof reagent-grade absolute ethyl alcohol
were purchased from Aldrich Chemical Co. (Milwaukee, WI).
Isopropyl alcohol (99.99%) was ordered from VWR (Oak
Ridge, TN). All water used in the study was doubly distilled
and deionized, so that the electrical conductivity was main-
tained below 18 mmho/cm. All solvents were used as received;
no further modifications or purification were performed.
Three different methods were used in this study for the syn-
thesis of ultrafine hydrous-zirconia particles:
1
4
(
I) Forced hydrolysis in aqueous inorganic salt solu-
sample is observed during experimentation. During a sample
test, powders were dispersed in methyl alcohol or ethyl alcohol
to form a slurry and then spread onto the top surface of the
Rh/Pt heating strip. After alcohol evaporation, a thin layer of
powder was uniformly distributed on the strip surface. To pre-
cisely maintain and measure the sample temperature (±1°C), a
surrounding heater was used around the heating strip and a
10% Rh/Pt thermocouple (type S) was welded to the bottom
surface of the thin heating strip. In addition, the high-
temperature furnace was provided with internal baffles to re-
duce thermal radiation directly to the water-cooled walls of the
furnace. The high-temperature chamber allowed for variation
in atmospheric composition; however, these experiments were
all conducted in flowing air at 1 atm. Data were taken by
heating the sample to predetermined temperatures (30°, 100°,
400°, 600°, 800°, 1000°, 1200°, 1400°C) at a heating rate of
0.5°C/min, and then cooling the sample to 30°C at a rate of
50°C/min. The holding time at each predetermined temperature
was 1 h before the XRD scan began. Data acquisition was
performed using DMS-NT software (Scintag Inc., Cupertino,
CA) and data analysis was undertaken using Jade software
(Materials Data Inc., Livermore, CA). Crystallite size was es-
timated from the broadening of the X-ray diffraction peaks,
using the Scherrer equation from the diffraction peak breadths
(full width at half-maximum).
tions: A 1.0M ZrOCl aqueous solution was usually prepared
2
as a stock solution and kept in a refrigerator at 4°C to avoid
possible hydrolytic aging at room temperature. (Note: In a
several-month-old solution, SAXS measurements indicated the
presence of zirconium species larger than the ∼ 4.5 Å radius of
gyration zirconium tetramers which exist in the fresh solution.)
Various concentrations (ranging from 0.01 to 2.0M dissolved
zirconium, typically 0.05M for preparation of powder charac-
terization sample) of ZrOCl aqueous solutions were then pre-
2
pared from the stock solution and incubated at 100°C for 3
days.
(II) Homogeneous precipitation in solutions of inorganic
salt dissolved in mixed water–alcohol solvents: The same
ZrOCl stock solution was used. In addition, isopropyl alcohol
2
was used as a dielectric-tuning agent for the alcohol–water
mixture solvent and hydroxyl propyl cellulose (HPC) as col-
loidal particle dispersant, which added steric forces between
particles. For this study, the mixture solution containing
−
3
3
ZrOCl (0.05 to 0.2M) and HPC (2.0 × 10 g/cm ) in mixed
2
alcohol–water solution (volume ratio of isopropyl alcohol to
water ס 5) was incubated at 100°C for several hours to days.
(
III) Alkoxide hydrolysis and condensation: Zirconium
tetra-n-butoxide (0.1 and 0.2M) solution in ethyl alcohol was
mixed at room temperature (22°C) with an equal volume of
water–ethyl alcohol solution (0.2, 0.4, and 0.6M). The water-
to-alkoxide molar concentration ratio was maintained at 2.0.
All solution incubations or agings were carried out in closely
capped glass vials or Nalgene plastic bottles. Milky white zir-
conia sol solutions obtained by the above procedures were
centrifuged and washed with deionized water several times and
freeze-dried.
III. Results and Discussion
The three different synthesis processes generated zirconia
particles with different characteristics. Figure 1 shows a rep-
resentative TEM photograph of zirconia particles produced by
the forced hydrolysis of 0.075M ZrOCl in aqueous solution.
2
The evolution and growth of hydrous-zirconia ultrafine sol
particles in liquid solutions were monitored with a custom-
designed dynamic light scattering (DLS) spectrometer. Before
the nucleation, early-stage polymeric cluster formation and
growth during the forced hydrolysis of zirconia tetramers were
measured by small-angle X-ray scattering (SAXS) at the Oak
Ridge National Laboratory (ORNL) user facility. A detailed
Near cubic-shaped, nanosize particles with a narrow size dis-
tribution (∼ 50 nm average size) were obtained. Each of the
particles was an aggregate of many needlelike, small primary
particles measuring ∼ 3 nm (rod diameter) by 5 nm (rod length).
This TEM observation on the shape of primary particles agrees
with our earlier SAXS studies. Depending on the initial dis-
solved Zr concentration (0.05 to 0.5M), particle sizes ranged
6

September 1999
Nanocrystallization and Phase Transformation in Zirconia Particles
2315
mechanisms are not clear. Nucleation occurs probably by either
hydrolytic reactions or a salting-out phenomenon. Growth
could be via diffusive molecular (i.e., tetramer) deposition or
an aggregation of primary particles. However, the effect of
isopropyl alcohol can be explained rather well by the dielectric
property of the water–isopropyl alcohol solvent. The presence
of the isopropyl alcohol reduced the dielectric constant of the
solution. Furthermore, elevation of the incubation temperature
also contributed to a reduction in the dielectric constant of the
mixed water–isopropyl alcohol solvent. The solubility of the
zirconium species in solution becomes lower when the dielec-
tric constant of the solvent is decreased. Particle formation in
the mixed solvent could occur at temperatures (such as 60°C)
that are too low for forced hydrolysis to take place. Therefore,
particle formation could be totally attributed to the salting-out
phenomenon due to the dielectric constant reduction, instead of
hydrolytic condensation. On the other hand, the presence of
isopropyl alcohol dramatically enhanced the particle formation
kinetics. Precipitation in a mixed water–alcohol system could
be completed in as little as 1 h, rather than days as required for
forced hydrolysis.
Fig. 1. Nanosize monoclinic zirconia particles with narrow size dis-
tribution produced from forced hydrolysis of aqueous solution of zir-
conyl chloride salt.
The sol–gel method using alkoxides as molecular precursors
has been a very popular approach for monodispersed ceramic
powder production. In this study, hydrous zirconia powder
samples were also prepared by this method by using optimal
conditions as reported in the literature (i.e., 0.1M zirconium
alkoxide, 0.2M water in ethanol). Powders produced by this
method were used only as a baseline product for comparison
with the powders obtained from the other two methods using
the inorganic salt precursor. Hydrolysis and condensation of
zirconium tetrabutoxide in ethanol–water solution produced
submicrometer-size microspheres, as shown in the SEM pho-
tograph (Fig. 4).
from tens to several hundred nanometers. Also, as the initial Zr
concentration was increased from 0.01 to 0.2M, the particle
size (Fig. 2) and hydrolysis reaction kinetics increased. In ad-
dition, the shape of the particle became more cubelike. How-
ever, increasing the zirconium concentration to greater than
1
1
0
.2M inhibited the kinetics because of the high acidity and
reduced the particle size. This observed effect of ZrOCl con-
2
centration on particle size is in agreement with the most recent
1
5
results obtained by Matsui and Ohgai. They determined that
+
H ion concentration produced during hydrolysis plays a sig-
The growth of hydrous zirconia sol particles in ZrOCl so-
2
nificant role in regulating the secondary particle size of hy-
drous zirconia. Detailed studies on nanoscale particle synthesis
by the forced hydrolysis of this system at low zirconium con-
lutions, in the presence and absence of isopropyl alcohol, was
monitored with the DLS spectrometer (Fig. 5). Note that the
early-stage data (when the particle size was less than 5 nm
during the induction period) for the forced hydrolysis process
centrations (<0.2M) has been extensively explored in our ear-
lier work.5,6
(
Fig. 5(a)) was measured with the SAXS instrument. It can be
In contrast, particles obtained from the homogeneous pre-
seen that the forced hydrolysis was a relatively slow process,
usually taking days to complete. Under equivalent conditions
cipitation of ZrOCl in water–isopropyl alcohol solutions are
2
perfect monodispersed microspheres, with sizes ranging from
nanometers to a few micrometers. The lower the initial zirco-
nium concentration in the mixed water–alcohol solvent, the
smaller the microsphere size. Typical SEM photographs are
shown in Fig. 3 The exact particle nucleation and growth
(
i.e., the same initial zirconium concentration and aging tem-
perature), the presence of the isopropyl alcohol shortened the
induction period dramatically and also increased the particle
growth rate. The shape of the growth curve shown in Fig. 5(b)
indicated a possible uncontrolled aggregation of microspheres
and further agglomeration of the aggregates. Such colloidal
coagulation eventually generated large agglomerates that were
beyond the size measuring capability of the DLS spectrometer.
The agglomerates must have been loosely bound and easily
dispersed because the SEM photograph showed individually
separated microspheres. On the other hand, we can see that the
forced hydrolysis method produced nanosize particles, whereas
the particles produced by the mixed solvent system were much
larger under the conditions shown in Fig. 5(b) (isopropyl al-
cohol was 50 vol%). However, the particle size in the mixed
solvent system is controllable, and smaller or even nanosize
monodispersed microsphere production can be achieved (Fig.
3
). Detailed results from extensive studies of homogeneous
precipitation in mixed solvent systems are reported elsewhere.
The postsynthesis characterization of powders prepared by
various methods was also a part of this study. First, BET analy-
sis of cube-shaped powders formed by forced hydrolysis (of
0
.05M ZrOCl , 100°C, 3 days), microsphere powders by
2
mixed-solvent precipitation (0.05M ZrOCl2 containing 50
vol% isopropyl alcohol, 100°C, 15 h), and the original ZrOCl2
salt precursor showed specific surface areas of approximately
2
2
2
3.75 m /g, <0.01 m /g, and negligible amounts, respectively.
Figure 6 shows typical nitrogen adsorption/desorption data for
cube-shaped zirconia powder formed by forced hydrolysis. We
observed that the freeze-drying process caused a significant
Fig. 2. Influence of initial ZrOCl salt concentration on the size of
final colloidal sol particles.
2

2316
Journal of the American Ceramic Society—Hu et al.
Vol. 82, No. 9
Fig. 3. Monodispersed zirconia microspheres by homogeneous precipitation of ZrOCl in mixed alcohol–water solutions: (a) submicrometer-size
2
spheres produced from 0.2M ZrOCl 50 vol% alcohol–water solutions incubated at 100°C for 2.5 h; (b) nanometer-size spheres produced from
2
0
.05M ZrOCl 50 vol% alcohol–water solutions incubated at 100°C for 2.5 h.
2
therms.” Various temperatures for the exotherm have been re-
1
7–20
ported around 430°, 400°, 420–440°, and 460°C.
The glow
exotherm was attributed to a rapid release of bound hydroxyl
1
,17,21–23
groups, which is a prerequisite for crystallization.
Oth-
ers think that the exotherm is due to the transformation of
zirconium oxyhydroxide to the amorphous oxide; that is,
2
0
ZrO(OH) → ZrO иH O. Most recently, two possible sepa-
2
2
2
rate transformation events were proposed to occur at the same
glow” temperature: (1) crystallization of amorphous materials
into small-size single-domain crystals and (2) exothermal
events commonly referred to as a glow phenomenon, wherein
small crystals coalesce into large particles containing one or
“
1
4
more grain boundaries. The sudden change in slope of the
DTA curves between 900° and 1200°C (Fig. 7(b)) is an indi-
cation of a phase transition, corresponding to monoclinic-to-
tetragonal transformation as shown later by XRD data (Fig. 8).
Nevertheless, the DTA curve for the cube-shaped particles pro-
duced by forced hydrolysis shows no obvious sharp peak at
50°C, only a broad curvature between 200° and 600°C.
It is well known that zirconia is a typical polymorphic ma-
terial which may exist in monoclinic, tetragonal, cubic, and
4
Fig. 4. Monodispersed zirconia microspheres produced by hydroly-
sis/condensation of zirconium tetra-n-butoxide.
2
4
orthorhombic crystalline forms at different temperatures. At
room temperature, zirconia has the monoclinic baddeleyite
structure, which is thermodynamically stable below 1150°C.
At a temperature of approximately 1150°C, the first displacive
martensitic phase transition from monoclinic to tetragonal
degree of particle aggregation and caking, which may be re-
sponsible for the low specific surface area of dried powders.
Preliminary studies on the crystallization and phase transitions
2
5,26
of the zirconia powders and ZrOCl salt were conducted by the
phases occurs.
At still higher temperature of approximately
2
TGA/DTA measurements (Fig. 7). It is clear that the micro-
spheres produced by mixed-solvent precipitation showed simi-
larities to the salt precursor in both TGA and DTA plots, except
that the salt seems to be more hydrated and thus lose more
weight during heating. In the TGA curve, the microsphere
powders show a total weight loss percentage similar to
2370°C, the tetragonal phase transforms into cubic phase with
a fluorite-type crystal structure which is stable up to the melt-
2
7,28
ing point (2680 ± 15°C).
However, the phase formation
and transformation behavior of hydrous-zirconia powders are
affected significantly by various synthesis conditions or pro-
cessing parameters such as pH, aging time, reaction tempera-
the theoretical weight loss value (22.61%) of Zr(OH) or
ture, precursor concentration, solvent type, and variation of
4
16
1,29–32
ZrO и2H O. Thus, the composition of the microspheres is
precursor salt
and, thus, may not always follow the tra-
2
2
thought to be similar to ZrO и2H O, which may result from
ditional phase transition route. The two zirconia powders and
the ZrOCl were further analyzed by HTXRD, a complemen-
2
2
combination of zirconyl tetrameric cations. There was a sudden
loss of weight around 450°C for both microspheres and salt.
The cube-shaped particles formed by forced hydrolysis were
the least hydrated and showed no sudden weight loss at 450°C.
In the DTA plot (Fig. 7(b)), the endothermic peak(s) and a
large weight loss appear between room temperature and 200°C,
corresponding to the thermal dehydration or evaporation of
physically adsorbed water. At about 450°C, a narrow, sharp
exothermic peak appeared for both microspheres and the salt
sample, indicating the crystallization of zirconia. Previous
studies have shown similar exothermic peaks in this tempera-
ture range (400–460°C) and are referred to as “glow exo-
2
tary technique to TGA/DTA. Figure 8 presents the XRD spec-
tra at various temperatures for cube-shaped zirconia powders
synthesized by forced hydrolysis. It can be seen that without
any heat treatment, the cube-shaped nanoparticles are mono-
clinic crystalline material (with crystallite sizes of approxi-
mately 14 nm at 30°C). Particularly noteworthy, the mono-
clinic peaks (at both 28.174° and 31.467° 2␪) gradually
increased in intensity when the sample was heated from 30°,
100°, 400°, 600°, 800°, up to 1000°C (1-h holding time at each
temperature), indicating further nucleation/growth of the
monoclinic nanocrystals inside the powder particles. Further

September 1999
Nanocrystallization and Phase Transformation in Zirconia Particles
2317
Fig. 5. Growth of zirconia sol particles in thermally incubated solutions of ZrOCl : (a) no alcohol solvent, forced hydrolysis; (b) with alcohol
2
solvent in solution, mixed-solvent precipitation.
Fig. 6. BET specific surface area analysis of cube-shaped, nanosize zirconia powders.
crystallization/growth of the monoclinic crystallites occurs un-
til their transition to the tetragonal phase (at 31.167° 2␪) around
above 400°C (Figs. 9 and 10). For both microspheres and the
salt the tetragonal phase appears first as the temperature is
increased. Then the monoclinic phase evolves at higher tem-
peratures and coexists with the tetragonal phase. The tetragonal
phase exists from the crystallization temperature (400°C for
microspheres and 600°C for the salt) all the way to 1400°C.
However, when the temperature is cooled down to 30°C or
room temperature, only monoclinic phase is present. Evidently,
during the solution synthesis the presence of the isopropyl
alcohol increases the particle growth rate by enhancing the
aggregation of zirconium species in the solution but inhibits the
ordered, hydrolytic arrangement of zirconium tetrameric spe-
cies and, thus, the monoclinic crystalline phase formation in the
microspheres. It is also worthwhile to note that the mixed-
solvent synthesis conditions (such as aging time, temperature,
alcohol-to-water volume ratio) affect the thermal crystalliza-
tion and phase-transition behavior of the microsphere powders.
We plan to report detailed studies of such effects later.
1
1
200°C. The tetragonal phase exists between 1200° and
400°C and monoclinic phase is completely transformed into
the tetragonal phase at 1400°C. This type of monoclinic-phase-
growth XRD data with increasing temperature is reported here
for the first time. When the temperature dropped from 1400°C
back to 30°C, the tetragonal phase was completely transformed
to the monoclinic phase, which is the stable phase at room
temperature. No metastable tetragonal phase stabilization was
observed. In addition, the percentage ratio of monoclinic phase
to tetragonal phase (m/t) is shown on the left-hand side of each
XRD spectrum in Fig. 8. (Note: Data are presented in the same
manner later in Figs. 9, 10, and 11.) This phase percentage ratio
was estimated from the integrated peak intensity according to
3
3
the method of Garvie and Nicholson.
In contrast to the cube-shaped particles, microspheres pro-
duced by mixed-solvent precipitation and ZrOCl salt precur-
2
sor are “amorphous” to X-rays if they are not heat-treated
As a baseline for comparison, the microsphere powders pre-

2318
Journal of the American Ceramic Society—Hu et al.
Vol. 82, No. 9
Fig. 7. Comparisons of TGA/DTA data of zirconia powders and precursor inorganic salt: (a) TGA plot, (b) DTA plot.
Fig. 8. HTXRD analysis of nanocrystalline cube-shaped zirconia
particles synthesized by forced hydrolysis.
Fig. 9. HTXRD analysis of ultrafine zirconia microspheres synthe-
sized by mixed-solvent precipitation.
pared by the alkoxide hydrolysis/condensation method also
tested by the HTXRD (Fig. 11) were found to be amorphous to
X-rays at room temperature. Only the tetragonal phase (no
monoclinic phase) appeared at temperatures from 600° to
ticles were heated from 400° to 1000°C in air. Such monoclinic
nanocrystal growth data agree very well with results reported
3
5
previously by Murase and Kato under a dry air or nitrogen
atmosphere (Fig. 12). In addition, Murase and Kato found that
the presence of water vapor during heat treatment stimulates
monoclinic zirconia crystal growth significantly. For the mi-
crospheres by mixed-solvent precipitation, the nanocrystal
growth of the metastable tetragonal phase as a function of
temperature (below 1200°C) is comparable with the experi-
1
400°C. As always, the tetragonal phase was reversed to the
monoclinic phase when the sample was cooled down from
400° to 30°C. The crystallization temperature for the micro-
1
sphere from mixed-solvent synthesis is around 400°C, which is
much lower than that of the salt (∼ 600°C) and the microspheres
from the alkoxide method. Such tetragonal-phase evolution
1
3
(corresponding to explosive crystallization or exotherm) from
mental crystallite size data previously reported by Garvie,
an amorphous material (i.e., hydrous zirconia gels) was re-
growing from 12.4 to 30 nm. In the case of heating ZrOCl2
hydrated salt, the salt crystal (believed to have a tetramer struc-
1
7
ported previously. Our current study is in agreement with
earlier work on the tetragonal crystal evolution from amor-
phous zirconia gel materials. HTXRD data for the amorphous
microsphere zirconia particles (Figs. 9 and 11) support the
belief that the formation of tetragonal zirconia is closely related
to the existence of amorphous zirconia because of their struc-
tural similarity. Topotatic crystallization on nuclei in the amor-
phous zirconia has been proposed as the mechanism of tetrag-
8 36
+
tural unit [Zr(OH) 4H O] )
becomes X-ray amorphous at
2
2
4
100° and 400°C. By further increasing the temperature, crystal-
lite size for the tetragonal phase grows from 12.3 to over 100 nm.
At temperatures higher than 1200°C, the effective crystallite
sizes of the tetragonal phase for all samples shown in Table I
are considerably larger (>100 nm) than the critical size (∼ 30
3
4,35,37–39
nm) previously reported for the metastable zirconia.
3
4
onal zirconia formation under hydrothermal conditions.
Garvie’s supporters believe that metastable tetragonal phase
can be stabilized only for those tetragonal crystals less than 30
nm. In other words, crystallites smaller than 30 nm could exist
in the tetragonal form while particles larger than this size
would tend to transform to the more stable monoclinic form.
Our 30°C data after cooling (Table I) clearly show that crys-
tallite size after the heating cycle is greater than the 30-nm
Table I shows the thermally induced nanocrystal evolution
and growth for the ZrOCl salt and the powders from the in-
2
organic salt. It is obvious that with increasing temperature, the
size of the crystallites increases. Nanocrystal growth depends
on the methods of zirconia preparation. Monoclinic crystallite
size increased from 9.3 to 41 nm when the cube-shaped par-

September 1999
Nanocrystallization and Phase Transformation in Zirconia Particles
2319
Fig. 10. HTXRD analysis of zirconyl chloride octahydrate salt.
Fig. 12. Correlation between nanocrystal growth of monoclinic zir-
conia and calcination temperature.
It is important to stress that for nanocrystalline cube-shaped
nanoparticles, the monoclinic nanocrystal growth and phase
transformation (monoclinic to tetragonal) that take place when
the temperature is increased from 30° to 1400°C in air (Fig. 8)
is an unusual but interesting case:
[
amorphous + monoclinic]→[monoclinic(crystallization)] →
[monoclinic + tetragonal)] → [pure tetragonal]
Traditionally, upon increasing calcination temperature, most
amorphous zirconia precursor materials convert to the tetrag-
onal phase first and then transform to monoclinic phase at
higher temperature (∼ 600°C); that is, tetragonal phase forms
preferentially to the monoclinic phase during crystallization of
1
,39–41
Fig. 11. HTXRD analysis of ziconia powders from hydrolysis/
condensation of zirconium tetra-n-butoxide in ethanol solution.
amorphous hydrous zirconia.
In the case of nanocrystal-
line zirconia precipitated from acidic salt solutions by hydro-
thermal treatment, an important exception to the traditional
7
,32,42
crystallization route occurs.
face energy theory,
According to Garvie’s sur-
the existence of monoclinic nanocrys-
talline zirconia would be attributed to a lowering of the surface
1
3,37
critical size. This may explain why in all of our samples no
metastable tetragonal phase remained stable when the tempera-
ture cooled down to room temperature. Several earlier works
reported that homogeneous precipitation produced zirconia
powders which contained single metastable tetragonal crystals
4
2
free energy under acidic conditions. The stability of the meta-
stable tetragonal phase has been attributed to the lower surface
energy of the tetragonal phase compared with the monoclinic
phase. The surface energy theory cannot explain entirely the
findings of this study, particularly why monoclinic crystal evo-
lution and growth occurred with increasing temperature, rather
than tetragonal crystal evolution. In fact, an earlier work by
Morgan reported phenomena that contradict the critical size
1
,4
exceeding 100 nm. They believed that the existence of the
large tetragonal monocrystals was due to the very fine internal
porosity or substructure of the closely oriented nanocrystallites
within the domains. The fine internal porosity gave rise to a
surface area/volume ratio sufficient to stabilize tetragonal zir-
conia by the same mechanism as in nanocrystalline powders.
4
2
theory. He obtained ultrafine zirconia particles as small as 6
Table I. Size Determination of Nanocrystals Evolved in Powders during Heat Treatment by HTXRD
Size (nm)
Cube-shaped zirconia powders
by forced hydrolysis
Microspheres of zirconia by
mixed-solvent precipitation
ZrOCl
2
salt precursor
Powders from alkoxide
hydrolysis/condensation
(as received)
Temperature
°C)
Monoclinic
phase
Tetragonal
phase
Monoclinic
phase
Tetragonal
phase
Monoclinic
phase
Tetragonal
phase
Monoclinic
phase
Tetragonal
phase
(
4
6
8
00
00
00
9.3
12.4
17.5
21.0
25.3
30.0
45.0
‡
11.0
17.5
ND
ND
12.3
11.6
33.5
38.5
19.5
41.0
21.0
1
1
1
000
200
400
25.5
33.0 to 49.0
30.0 to 45.0
90.0 to 100.0
42.8
40.0
40.0 to 64.0
†
>100.0
>100.0
>100.0
>100.0
>100.0
>100.0
30
60.0 to 100.0
44.0 to 48.0
60.5 to 68.0
52.5
†
‡
Peak (or crystalline phase) does not exist. Cannot be accurately determined from the XRD data.

2320
Journal of the American Ceramic Society—Hu et al.
Vol. 82, No. 9
13R. C. Garvie, “The Occurrence of Metastable Tetragonal Zirconia as a
nm in the monoclinic phase. Based on TEM particle size and
XRD crystallite size, Srinivasan et al. also concluded that the
tetragonal-to-monoclinic transformation does not appear to be
Crystallite Size Effect,” J. Phys. Chem., 69 [4] 1238–43 (1965).
14
R. Srinavisan and B. H. Davis, “Glow Phenomenon and Crystallization:
Evidence That They Are Different Events for Hafnium–Zirconium Mixed Ox-
4
3
due to a critical size effect. When the primary particles ag-
gregate together with the formation of necks, D e´ champs et al.
proposed a branched-chain model in which the tetragonal-to-
monoclinic transformation in zirconia is independent of the
ides,” J. Colloid Interface Sci., 156, 400–405 (1993).
15
K. Matsui and M. Ohgai, “Formation Mechanism of Hydrous-Zirconia
Produced by Hydrolysis of ZrOCl2 Solutions,” J. Am. Ceram. Soc., 80 [8]
1
949–56 (1997).
16Y. T. Moon, H. K. Park, D. K. Kim, and C. H. Kim, “Preparation of Mono-
4
size of the aggregate or domain.
disperse and Spherical Zirconia Powders by Heating of Alcohol–Aqueous Salt
Solutions,” J. Am. Ceram. Soc., 78 [10] 2690–94 (1995).
17
J. Livage, K. Doi, and C. Mazieres, “Nature and Thermal Evolution of
Amorphous Hydrated Zirconium Oxide,” J. Am. Ceram. Soc., 51, 349–53
1968).
18E. D. Whitney, “Observations on the Nature of Hydrous Zirconia,” J. Am.
Ceram. Soc., 53 [12] 697–98 (1970).
IV. Summary
(
Inorganic salt can be used as a good precursor in the pro-
duction of monodispersed ultrafine ceramic powders and pro-
vides the same level of quality as that obtained using higher-
cost alkoxide precursor. Ultrafine monodispersed zirconia
particles were successfully synthesized by three different meth-
ods for comparative studies. The particle-synthesis methods
19
M. T. Torrolvo, M. A. Alario, and J. Soria, “Crystallization Behavior of
Zirconium Oxide Gels,” J. Catal., 86, 473–76 (1984).
20
S. A. Selim and T. M. E. Akkard, “Thermal Decomposition, Pore Structure
and Heats of Immersion of Zirconia Gel,” J. Appl. Chem. Biotechnol., 27, 58–66
(1977).
21G. Gimblet, A. A. Rahman, and K. S. W. Sing, “Thermal and Related
Studies of Some Zirconia Gels,” J. Chem. Technol. Biotechnol., 30, 51–64
(1980).
were (1) forced hydrolysis of ZrOCl in aqueous solutions, (2)
2
mixed-solvent precipitation of ZrOCl by adjusting the dielec-
2
22P. D. L. Mercera, J. G. Van Ommen, E. B. M. Doesburg, A. J. Burggraaf,
and J. R. H. Roos, “Zirconia as a Support for Catalysts—Evolution of the Tex-
ture and Structure on Calcination in Air,” Appl. Catal., 57, 127–48 (1990).
tric constant of the mixture solutions, and (3) hydrolysis/
condensation of zirconium alkoxide. For the forced hydrolysis
and mixed-solvent precipitation processes, SAXS and DLS
were effective tools used to monitor the particle nucleation and
growth of the sol particles. The forced hydrolysis is a slow
process that produced monoclinic, cube-shaped nanoparticles
of zirconia. In contrast, the mixed-solvent precipitation and the
alkoxide method produced amorphous microspheres. Produc-
tion of nanosize (<100 nm) microspheres can be achieved in
the mixed-solvent precipitation of inorganic salt solutions.
Powder characterizations indicated that the synthesis methods
or processing conditions affect the behavior of crystallization
and phase transition, which does not follow the traditional
phase transformation route for zirconia. It was discovered that
the cube-shaped particles produced from forced hydrolysis fol-
low a unique nanocrystal evolution and growth pathway; that
is, monoclinic crystals can evolve and grow before the appear-
ance of tetragonal crystals during the heating process. The
microspheres from mixed-solvent synthesis exhibit a mixture
of tetragonal and monoclinical phases, while microspheres
from alkoxide hydrolysis/condensation show a pure tetragonal-
phase formation.
23
R. Srinivasan, M. B. Harris, S. F. Simpson, R. J. Deangelis, and B. H.
Davis, “Zirconium Oxide Crystal Phase: The Role of the pH and Time to Attain
the Final pH for Precipitation of the Hydrous Oxide,” J. Mater. Res., 3, 787–97
(1988).
24G. Fisher, “Zirconia: Ceramic Engineering’s Toughness Challenge,” Am.
Ceram. Soc. Bull., 65 [10] 1355–60 (1986).
25
E. Ryshkewitch and D. W. Richerson, Oxide Ceramics, Physical Chemistry
and Technology, 2nd ed. General Ceramics, Haskell, NJ, 1985.
26
C. J. Howard, R. J. Hill, and B. E. Reichert, “Structures of the ZrO Poly-
2
morphs at Room Temperature by High-Resolution Neutron Powder Diffrac-
tion,” Acta Crystallogr., Sect. B, 44, 116–20 (1988).
27
P. Aldebert and J.-P. Traverse, “Structure and Ionic Mobility of Zirconia at
High Temperature,” J. Am. Ceram. Soc., 68, 34–40 (1985).
28
D. K. Smith and C. F. Cline, “Verification of Existence of Cubic Zirconia
at High Temperature,” J. Am. Ceram. Soc., 45, 249–50 (1962).
29
X. Bokhimi, A. Morales, O. Novaro, M. Portilla, T. L o´ pez, F. Tzompantzi,
and R. G o´ mez, “Tetragonal Nanophase Stabilization in Nondoped Sol–Gel Zir-
conia Prepared with Different Hydrolysis Catalysts,” J. Solid State Chem., 135,
28–35 (1998).
30M. C. Silva, G. Trolloard, O. Masson, R. Guinebretiere, A. Dauger, A.
Lecomte, and B. Frit, “Early Stages of Crystallization in Gel Derived ZrO2
Precursors,” J. Sol–Gel Sci. Technol., 8, 419–24 (1997).
31
M. L. Rojas-Cervantes, R. M. Martin-Aranda, A. J. Lopez-Peinando, and J.
De D. Lopez-Gonzalez, “ZrO Obtained by the Sol–Gel Method: Influence of
Synthesis Parameters on Physical and Structural Characteristics,” J. Mater. Sci.,
2
2
9, 3743–48 (1994).
32R. P. Denkewicz, Jr., K. S. TenHuisen, and J. H. Adair, “Hydrothermal
References
1B. Djuri cˇ i c´ , S. Pickering, D. McGarry, P. Glaude, P. Tambuyser, and K.
Schuster, “The Properties of Zirconia Powders Produced by Homogeneous Pre-
cipitation,” Ceram. Int., 21, 195–206 (1995).
Crystallization Kinetics of m-ZrO and t-ZrO ,” J. Mater. Res., 5 [11] 2698–705
2 2
(1990).
33R. C. Garvie and P. S. Nicholson, “Phase Analysis in Zirconia Systems,” J.
Am. Ceram. Soc., 55, 303 (1972)
2
R. H. J. Hannink and M. V. Swain, “Progress in Transformation Toughening
34
of Ceramics,” Annu. Rev. Mater. Sci., 24, 359–408 (1994).
E. Tani, M. Yoshimura, and S. Somia, “Formation of Ultrafine Tetragonal
3
A. H. Heuer and L. W. Hobbs (Eds.), Advances in Ceramics, Vol. 3, Science
ZrO Powder under Hydrothermal Conditions,” J. Am. Ceram. Soc., 66, 11–14
2
and Technology of Zirconia. American Ceramic Society, Columbus, OH, 1981.
(1983).
4
35Y. Murase and E. Kato, “Role of Water Vapour in Crystallite Growth and
M. D e´ champs, B. Djuri cˇ i c´ , and S. Pickering, “Structure of Zirconia Prepared
by Homogeneous Precipitation,” J. Am. Ceram. Soc., 78, 2873–80 (1995).
Tetragonal–Monoclinic Phase Transformation of ZrO ,” J. Am. Ceram. Soc.,
2
5
M. Z.-C. Hu, M. T. Harris, and C. H. Byers, “Nucleation and Growth Ki-
66, 196–200 (1983).
netics for Synthesis of Nanometric Zirconia Particles by Forced Hydrolysis,” J.
Colloid Interface Sci., 198, 87–99 (1998).
36T. C. Mak, “Refinement of the Crystal Structure of Zirconyl Chloride Oc-
tahydrate,” Can. J. Chem., 46, 3491–97 (1968).
6
37
M. Z.-C. Hu, J. T. Zielke, J.-S. Lin, and C. H. Byers, “Small-Angle X-ray
R. C. Garvie, “Stablization of the Tetragonal Structure in Zirconia Micro-
Scattering Studies of Early-Stage Colloids Formation by Thermohydrolytic Po-
crystals,” J. Phys. Chem., 82, 218–24 (1978).
38
lymerization of Aqueous Salt Solutions,” J. Mater. Res., 14, 103–13 (1999).
T. Mituhashi, M. Ichihara, and U. Tatsuke, “Characterization and Stabili-
zation of Metastable Tetragonal ZrO ,” J. Am. Ceram. Soc., 57, 97–101 (1974).
7
A. Bleier and R. M. Cannon, “Nucleation and Growth of Uniform m-ZrO ,”
2
2
39
Mater. Res. Soc. Symp. Proc., 73, 71–78 (1986).
R. Srinivasan, B. H. Davis, O. Burlcavin, and C. R. Hubbard, “Crystalli-
8
M. Chatry, M. Henry, and J. Livage, “Synthesis of Nonaggregated Nano-
zation and Phase Transformation Process in Zirconia: An In-Situ High-
Temperature X-ray Diffraction Study,” J. Am. Ceram. Soc., 75, 1217–22 (1992).
metric Crystalline Zirconia Particles,” Mater. Res. Bull., 29, 517–22 (1994).
9
40
H. Kumazawa, T. Inoue, and E. Sada, “Synthesis of Fine Spherical Zirconia
M. Yashima, T. Kato, M. Kakihana, M. A. Gulgun, Y. Matsuo, and M.
Particles by Controlled Hydrolysis of Zirconium Alkoxide in the High Tem-
Yoshimura, “Crystallization of Hafnia and Zirconia During the Pyrolysis of
perature Range Above 50°C,” Chem. Eng. J., 55, 93–96 (1994).
Acetate Gels,” J. Mater. Res., 12, 2575–83 (1997).
1
0
41
H. Kumazawa, Y. Hori, and E. Sada, “Synthesis of Spherical Zirconia Fine
G. T. Mamott, P. Barnes, S. E. Tarling, S. L. Jones, and C. J. Norman, “A
Particles by Controlled Hydrolysis of Zirconium Tetrabutoxide in 1-Propanol,”
Dynamic High Temperature XRPD Study of the Calcination of Zirconium Hy-
Chem. Eng. J., 51, 129–33 (1993).
droxide,” Powder Diffr., 3 [4] 234 (1988).
1
1
42
S.-K. Lee, I. Masaki, and M. Nobuyasu, “Influence of Alcoholic Solvent on
P. E. D. Morgan, “Synthesis of 6 nm Ultrafine Monoclinic Zirconia,” J.
Formation of Monodispersed Particles by Hydrolysis of Zirconia Tetra-n-
butoxide,” J. Ceram. Soc. Jpn., 99, 290–94 (1990).
T. Ogihara, N. Mitzutani, and M. Kato, “Growth Mechanism of Monodis-
persed ZrO2 Particles,” J. Am. Ceram. Soc., 72 [3] 421–26 (1989).
Am. Ceram. Soc., 67 [10] C-204–C-205 (1984).
43
R. Srinivasan, L. Rice, and B. H. Davis, “Critical Particle Size and Phase
Transformation in Zirconia: Transmission Electron Microscopy and X-ray Dif-
fraction Studies,” J. Am. Ceram. Soc., 73 [11] 3528–30 (1990).
1
2
ᮀ