High yield hydrothermal synthesis of nano-scale zirconia and YTZP

Article abstract of DOI:10.1111/j.1551-2916.2011.04806.x

Many techniques exist to synthesize zirconia and YTZP with a nano-scale particle size, but a recently developed hydrothermal precipitation procedure has shown promise in producing high-yield quantities of well-dispersed 8-10 nm zirconia and YTZP. Using bicine as a complexing agent, a homogeneous solution of zirconyl nitrate and yttrium nitrate is formed; after hydrothermal treatment, 8-10 nm tetragonal zirconia or YTZP particles are precipitated. With the modified hydrothermal procedure, powder yield was increased to 100 g of powder per liter of reactant solution without significantly changing the powder characteristics. X-ray diffraction and Raman spectroscopy were used to verify that the tetragonal phase was formed in the 1YTZP composition. Complementary particle size measurements were made using dynamic light scattering, BET nitrogen adsorption, and X-ray diffraction; differences in measured particle size are discussed, and all three techniques verify the 8-10 nm particle diameter. Particle morphology and impurity content were measured with transmission electron spectroscopy and X-ray fluorescence, respectively. To overcome the increased likelihood of agglomeration that accompanies the increased synthesis, wash solutions of deionized water, bicine, and oxalic acid were compared, and the particle-particle interactions of nano-YTZP in these wash solutions and in ethanol are modeled with modified DLVO theory. The results show that yttria leaching is prevented and dispersion is promoted by laundering with a solution of oxalic acid and dispersing into deionized water, both adjusted to a pH value of 9.

Full text of DOI:10.1111/j.1551-2916.2011.04806.x

J. Am. Ceram. Soc., 94 [12] 4239–4246 (2011)
DOI: 10.1111/j.1551-2916.2011.04806.x
© 2011 The American Ceramic Society
ournal
J
High Yield Hydrothermal Synthesis of Nano-Scale Zirconia and YTZP
Christopher J. Szepesi† and James H. Adair
Department of Materials Science and Engineering, College of Earth and Mineral Sciences,
The Pennsylvania State University, University Park, PA 16802
Many techniques exist to synthesize zirconia and YTZP with a
nano-scale particle size, but a recently developed hydrothermal
precipitation procedure has shown promise in producing high-
yield quantities of well-dispersed 8–10 nm zirconia and YTZP.
Using bicine as a complexing agent, a homogeneous solution of
zirconyl nitrate and yttrium nitrate is formed; after hydrother-
mal treatment, 8–10 nm tetragonal zirconia or YTZP particles
are precipitated. With the modified hydrothermal procedure,
powder yield was increased to 100 g of powder per liter of
reactant solution without significantly changing the powder
characteristics. X-ray diffraction and Raman spectroscopy were
used to verify that the tetragonal phase was formed in the
1YTZP composition. Complementary particle size measure-
ments were made using dynamic light scattering, BET nitrogen
adsorption, and X-ray diffraction; differences in measured par-
ticle size are discussed, and all three techniques verify the
8–10 nm particle diameter. Particle morphology and impurity
content were measured with transmission electron spectroscopy
and X-ray fluorescence, respectively. To overcome the
increased likelihood of agglomeration that accompanies the
increased synthesis, wash solutions of deionized water, bicine,
and oxalic acid were compared, and the particle–particle inter-
actions of nano-YTZP in these wash solutions and in ethanol
are modeled with modified DLVO theory. The results show
that yttria leaching is prevented and dispersion is promoted by
laundering with a solution of oxalic acid and dispersing into
deionized water, both adjusted to a pH value of 9.
suspensions of 8–10 nm zirconia or yttrium-doped tetragonal
zirconia polycrystals (YTZP) are hydrothermally prepared by
taking advantage of the complexation chemistry of the system.
Before precipitation, the participating metal ions are kept
in a dissolved state by a complexing agent to produce a
homogeneous solution. Precipitation takes place when the
complex is broken and large amounts of metal ions are
released into the solution. When this occurs, the supersatura-
tion of the solution is extremely high, and almost all of the
energy available for growth is consumed in the formation of
nuclei. As a consequence, little growth of the nuclei occurs
and the resulting particles are ultra-fine in size.
The choice of complexing agent in Kimel and Adair’s18
work was based on several considerations. First, the com-
plexing agent needed to form a stable complex with the metal
ions over a wide range of pH values. Second, the complexing
agent should impart charge to the surface of the newly
formed particles to promote dispersion in suspension after
precipitation. From previous studies19–21 and experimental
verification, bicine (N,N-bis(2-hydroxyethyl)glycine) was
selected as the appropriate complexing agent.
In the development of the technique by Kimel and Adair,
the concentration of zirconium in solution produced a relatively
small mass (~25 g) of nano-YTZP per liter of stock solution.
For this to be an economical process, the yield of nano-scale
material should be as high as possible and practical.
The scale-up of a precipitation procedure is often accom-
a greater tendency for agglomeration. The
panied by
increased yield of particulate also generates increased solids
loading, particle number concentrations, and ionic strength
in the precipitant solution. According to conventional DLVO
theory (named for authors Derjaguin and Landau,22 Verwey
and Overbeek23), which models attractive and repulsive inter-
actions between particles in suspension, increased solution
ionic strength leads to double layer compression and lower
energy barriers to agglomeration. The double layer thickness,
I. Introduction
ANO-SCALE zirconia materials offer unique mechanical
properties, such as enhanced transformation toughen-
N
ing with less dopant, and superplastic deformation behav-
ior.1–3 Several authors have demonstrated that high-density,
nano-grain ceramics can be fabricated with nano-scale zirco-
nia materials.4,5 However, authors have expressed skepticism
in whether nano-scale materials could be synthesized in
industrial quantities.6 For example, many popular chemical-
vapor synthesis methods7,8 require significant time and
energy to produce relatively small amounts of material. How-
ever, the perception that nano-scale materials are only practi-
cal for small-scale research purposes is becoming discarded
thanks, at least in part, to recent developments in wet-chemi-
cal precipitation,9–11 sol–gel synthesis,12–14 and hydrothermal
techniques.15–17 In fact, several grades of nano-zirconia mate-
rials are commercially available from a variety of sources.
One synthesis technique in particular, developed by Kimel
and Adair,18 is of particular interest for manufacturing larger
quantities of nano-zirconia. With this procedure, stable aqueous
or Debye–Huckel length (jꢀ1, m) is a characteristic measure-
¨
ment of the distance that the counterion cloud extends into
the suspension, and can be calculated with the equation24
1
2
ꢀ
ꢁ
8000pe2NAI
e0kdkBT
jꢀ1
¼
(1)
where e is the charge of an electron, NA is Avogadro’s number,
I is the suspension ionic strength in mol/L, e0 is the permittiv-
ity of free space, kd is the relative dielectric constant of the
medium, kB is Boltzmann’s constant, and T is absolute temper-
ature. Thus the Debye–Huckel length diminishes with Iꢀ1/2
.
a fourfold
For the synthesis procedure in question,
increase in yield equates to a fourfold increase in ionic
strength, calculated by
H. Chan—contributing editor
X
where zi is the valence of a dissolved species i, and ci is the
1
I ¼
Zi2ci
(2)
2
i
Manuscript No. 29389. Received March 01, 2011; approved July 22, 2011.
†Author to whom correspondence should be addressed. e-mail: christopher.szepesi@
am.umicore.com
molar concentration. Referring to Eq. (1), this increase in
4239

4240
Journal of the American Ceramic Society—Szepesi et al.
Vol. 94, No. 12
ionic strength compresses the double layer by a factor of 2,
which can be significant on the length scale of nano-particles.
For successful processing into green bodies and, ultimately,
high-density ceramics, the increased probability of agglomer-
ation must be accounted for with a washing procedure that
avoids the increased likelihood of agglomeration.
The usefulness of the precipitation process detailed in this
work, however, is that yields are only limited by the solubili-
ties of the reagents used. Upon precipitation, particles are
surrounded by a protective colloid layer, which acts as an
effective barrier to irreversible agglomeration. The process
also allows flexibility in the concentration of dopant incorpo-
rated into the material.
In the current work, nano-YTZP yields have been
increased to as high as 100 g/L (10 wt%) by increasing the
metal ion and bicine concentrations (along with appropriate
pH value compensation) without changing the product prop-
erties and without irreversible agglomeration. The effect of
washing solution additives on particle dispersion is also pre-
sented, and an effective washing and dispersion procedure is
identified, which minimizes agglomeration during laundering
despite the increased likelihood in the high-yield formulation.
After agitation, the mixture was placed into a modified
Parr hydrothermal vessel with a Teflon lining, described in
more detail elsewhere.18 The mixture was treated for 8 h at
200°C and cooled to room temperature.
To remove reaction byproducts, a displacement washing
procedure was employed. Three wash solutions were chosen
for this study: 0.01M oxalic acid(aq) (A.C.S. grade, Fisher Sci-
entific, Hanover Park, IL), 0.01M bicine(aq) (minimum 99%
titration, Sigma Aldrich), and deionized water, each adjusted
to pH 9 with tetraethylammonium hydroxide (20 w/w solution
in water, Sigma Aldrich) to minimize yttrium dissolution.25–27
The rationale behind the choice of these particular reagents
will be discussed later in this article. All laundering was carried
out in Nalgene containers to minimize contamination from
glassware at elevated pH values. All suspensions were washed
until a constant conductivity was achieved. In the final washing
(dispersion) step, the solids loading was adjusted to 20 wt% by
adding an appropriate amount of dispersing solution.
Dispersion was evaluated in washed suspensions using
dynamic light scattering (DLS) to measure particle size distri-
bution (Malvern ZetaSizer, Model Nano S; Malvern Instru-
ments, Worcestershire, U.K.) and zeta potential (ZetaPALS
Zeta Potential Analyzer, Brookhaven Instrument Corpora-
tion, Holtsville, NY). Suspensions were diluted to optical
transparency into pH-adjusted deionized water or ethanol for
these measurements. Particle interaction energies and suspen-
sion stability were calculated in the STABIL© v4.5 computer
program24 based on a modified DLVO theory used by Kimel
and Adair. The Hamaker constant used for tetragonal YTZP
was 8.8 9 10ꢀ20 J,28 and the approximation for small values
of ja (ratio of particle size to double layer thickness <<1)
was used in the calculation. The size of the adsorbed layers
for the oxalate and bicine molecules was the same as in Ki-
mel and Adair (0.37 and 0.75 nm, respectively). Suspension
stability is further characterized by the agglomeration proba-
bility, a, described by the equation24
II. Materials and Methods
Figure 1 presents a schematic flow diagram of the synthesis
procedure. Please note that the powder recovery and green
forming steps are outside the scope of this report. Nano-scale
zirconia and YTZP were prepared using zirconyl nitrate
hydrate and yttrium nitrate hexahydrate (Sigma-Aldrich,
Milwaukee, WI) as the sources for zirconium and yttrium
ions, respectively, with bicine (>99% titration, Sigma-
Aldrich) added to complex the ions in aqueous solution. The
molar ratio of bicine to Zr4+ was kept at 2:1.18 Typical con-
centrations of zirconyl nitrate were on the order of 0.5–1.0M
with yttrium nitrate added to produce between 1 and 4 mol%
yttria doping. Appropriate amounts of CO2-free deionized
water, bicine, zirconyl nitrate, and yttrium nitrate were added
to solid tetramethylammonium hydroxide pentahydrate
(TMAOH, >97%, Sigma-Aldrich) in a Nalgene bottle. The
mixture was shaken overnight, forming a translucent white
solution. The water contained in the TMAOH, when released
during agitation, acted as a medium in which the other
reagents could dissolve. The pH of the mixture was ꢁ 13.
R
1
1
expðVAðdÞ=kB
TÞ ds
TÞ ds
s2
expðVTðdÞ=kB
s2
2
a ¼
(3)
R
2
This value is also calculated with the Stabil© software,24
where VA(d) is the attractive energy curve, VT(d) is the total
energy curve, and s, the homologous separation distance, is
a1 þ a2 þ d
s ¼
(4)
a1 þ a2
ai is the diameter of particle i, and d is the separation
between overlapping double layers. When employing these
models, it was assumed that particles were spherical, mole-
cularly smooth, with charges evenly distributed over the
particle surface.
The product was further characterized through measure-
ment of specific surface area by N2 adsorption (Gemini II 2370
Surface Area Analyzer, Micromeritics Instrument Corpora-
tion, Norcross, GA), crystal structure and crystallite size by
X-ray diffraction (PAD V X-ray Diffractometer, Scintag, Inc.,
Sunnyvale, CA), phase determination by Raman spectroscopy
(Model PI-ECL-785-300-FC Raman Spectrometer, InPhoton-
ics, Inc., Norwood, MA), and compositional analysis by X-ray
fluorescence (Philips 1600/10 Wavelength Dispersive XRF,
courtesy of the Mineral Lab, Lakewood, CO). Crystallite size
was estimated from peak broadening of the (011) reflection in
the XRD pattern of dried powder with the Scherrer equation,
Kk
dXRD
¼
(5)
D
2h cos h
where K is a shape factor (usually equal to 0.9), and D2h is
the peak width in radians, and h is the diffraction angle.29
Fig. 1. A modified procedure to increase yield of hydrothermally
synthesized 8–10 nm YTZP is illustrated in this flow chart.

December 2011
Hydrothermal Synthesis of Nano-Scale Zirconia and YTZP
4241
The product was also observed using a transmission electron
microscope (Philips 420T TEM; FEI Company, Hillsboro,
OR) in the electron microscopy facility of the Materials
Research Institute at the Pennsylvania State University.
Average agglomeration number (AAN), which estimates
the average number of primary particles in an agglomerate,
was calculated as the ratio of the cube of the mean particle
size measured by light scattering (hydrodynamic diameter,
DDLS) to the cube of the equivalent spherical diameter mea-
sured by BET gas adsorption times a packing fraction,
assuming spherical particles.30 The equivalent spherical diam-
eter is given by
the higher yield of zirconia, it was uncertain what wash pro-
cedure was most appropriate.
Figure 2 illustrates an example of the particle size distribu-
tions of several washed suspensions as measured by dynamic
light scattering. Note that this particle sizing technique gives
a measurement of hydrodynamic diameter, DDLS, which is
affected by light scattering from the double layer. To confirm
the primary particle size, BET surface area and XRD peak
broadening were also used. Surface area values were mea-
sured to be between 100 and 120 m2/g, which corresponds to
a dBET between 8 and 10 nm using Eq. (6).
Equation (6) assumes that particles are spherical in shape,
which is a reasonable assumption, given the particles shown
in the TEM micrographs in Fig. 3, which shows undoped
zirconia and 1YTZP particles washed with oxalic acid solu-
tion. Crystallite size calculations from XRD peak broaden-
ing, shown in Fig. 4 for a 1YTZP sample, using the Scherrer
equation, result in values similar to those calculated from
surface area measurements. Therefore, from surface area
measurements, XRD analysis and TEM observations, it is
concluded that particles have an average diameter between 8
and 10 nm.
6
dBET
¼
(6)
ðSSA ꢂ qÞ
where dBET is an equivalent spherical diameter (lm), SSA is
the specific surface area (m2/g), and ρ is the density of the
material (g/cm3). The equivalent spherical diameter was con-
verted from a surface area basis to a volume basis, D, for
comparison with the volume-based hydrodynamic diameter
with the equation30
However, these calculated values are in slight disagreement
with the hydrodynamic diameter determined by light scatter-
ing. As particles are in a suspended, charged state during
measurements with this technique, a slight shift in the posi-
tion of the primary particle mode, as illustrated in Fig. 2 with
different wash solutions, suggests a difference in double layer
thickness with wash solution.32,33 In fact, the particle size
measured by light scattering being larger than that measured
by BET surface area, XRD, and TEM, verifies that the dou-
ble layer thickness and adsorbed species affect the particle
size measured by the light scattering technique.34,35 This effect
and corrections for it are discussed in the following section.
The XRD measurements were also ambiguous in differen-
tiating between the cubic and tetragonal phases of zirconia in
the 1YTZP product. Because of crystallite size effects, peak
broadening obscures any splitting associated with the tetrag-
onal phase. Therefore, to conclusively determine which phase
was produced, Raman spectra were taken on samples of
dry-pressed powder. Figure 5 illustrates such a spectrum of a
1YTZP product, which exhibits the six vibrational modes
associated with the tetragonal phase, not the single mode
associated with the cubic phase.36 A seventh mode was
attributed to residual organics.
ꢀ
ꢁ
6
4:6
2
logðDÞ ¼ log
þ
(7)
qSSA
sw
where sw is the width of the particle size distribution, usually
taken as 4 or 5 for a log-normally distributed powder.
III. Results and Discussion
(1) Synthesis Procedure and Characterization
Treatment time, reaction pH, and temperature were kept the
same as optimized in the Kimel and Adair procedure18 (8 h
at 200°C in a Teflon-lined hydrothermal vessel). These condi-
tions were chosen to hydrothermally decompose the bicine–
zirconium complex and precipitate the tetragonal phase.
Adair26 and Denkewicz et al.31 have generated solubility dia-
grams for the Zr–H2O and Y–H2O systems and have shown
that hydrothermal precipitation at pH values above 12 is
necessary to produce the stabilized tetragonal phase.
Alkaline metal bases are normally avoided in achieving
such a high pH because of possible deleterious effects during
subsequent processing, such as sintering. Consequently, a
concentrated organic base needed to be identified to adjust
solution pH without drastically increasing solution volume.
XRF was used to determine impurity concentrations and
to verify the level of yttrium doping within synthesized
particles. Typical impurities are given in Table I and consist
Solid
tetramethylammonium
hydroxide
pentahydrate
(TMAOH, >97%, Sigma-Aldrich, Milwaukee, WI) is readily
available, and, when mixed as a solid with the solid acidic
metal ion salts, forms a highly alkaline aqueous solution by
the release of water contained within the reagent structures.
Excess water can then be added to adjust the metal ion con-
centrations to the desired levels. Treatment time and temper-
ature were kept the same as optimized in the Kimel and
Adair procedure18 (8 h at 200°C in an unstirred Teflon-lined
hydrothermal vessel).
Zirconium concentration within the stock solutions was
limited to about 0.8M due to the volume increase associated
with higher reagent concentrations. Yields from these stock
solutions routinely achieve 100 g/L of nano-scale YTZP
upon hydrothermal treatment using the modified procedure
described in Fig. 1.
An often overlooked, but critical element in the solution
synthesis of nano-particulates is the laundering and recovery
steps to remove spectator ions while maintaining a well-dis-
persed system. To remove reaction byproducts, several wash-
ing steps are typically required. This is done with a standard
displacement washing procedure in which solids are concen-
trated from suspension by centrifugation, supernatant is
decanted and replaced with fresh wash solution, and solids
are redispersed with moderate ultrasonication. However, for
As-Washed 1YTZP (Zr-781)
Bicine
Bicine/Ethanol
Oxalate
Oxalate/Ethanol
Water
Water/Ethanol
14
12
10
8
6
4
2
0
10
100
Particle Size (nm)
Fig. 2. Hydrodynamic diameter measurements by dynamic light
scattering (DLS) show good dispersion in 10 w/w YTZP
suspensions. Differences in primary mode may be attributed to
variations in double layer thickness and hydrodynamic diameter with
wash solution additive.

4242
Journal of the American Ceramic Society—Szepesi et al.
Vol. 94, No. 12
(a)
(b)
Fig. 3. TEM micrographs of (a) undoped zirconia and (b) 1YTZP, both washed with oxalic acid solution, confirm particle sizes measured by
BET surface area and XRD peak broadening as between 8 and 10 nm.
Fig. 4. XRD analysis reveals that either the cubic or tetragonal phase is present in as-synthesized 1YTZP. Calculation of particle size from
peak broadening using the Scherrer equation yields crystallite sizes between 8 and 10 nm.
of low levels of sodium, magnesium, silicon, phosphorus, and
titanium.
needed with conventional dry forming techniques.38–40
Uncontrolled agglomeration can also be prevented because
wet processing routes use the nano-particles in a dispersed
state. Traditional dry-powder green forming techniques,
widely employed in industry, have also been shown to be
useful with nano-scale materials.41–44 However, aggregation
and poor powder flow characteristics, resulting from a large
number of particle–particle contacts and high internal fric-
tion, routinely result in inhomogeneities that can compromise
both the green and final microstructures.17,42,45–47 To recover
a material suitable for further processing, either wet or dry,
particles should first be well dispersed in a medium free of
reaction by-products and spectator ions. With this in mind,
wash solution additives were chosen to promote dispersion
as well as prevent the loss of yttrium dopant from synthe-
sized particles.
A yttrium concentration between 1.5 and 3 mol% is a rou-
tine target composition to produce transformation toughening
behavior in the final ceramic body.2,37 With their procedure,
Kimel and Adair18 were able to produce zirconia containing
1.7 mol% yttria. XRF measurements for batches prepared
with a 1.5 mol% yttria initial input indicate a final concentra-
tion of between 0.8 and 1 mol%. An advantage of this syn-
thesis technique, however, is that it allows flexibility in
composition by simply adjusting the beginning recipe.
(2) Washing and Dispersion
Several previous studies have demonstrated that wet process-
ing routes provide the best possible mechanism for obtaining
“bulk” nano-crystalline ceramics because wet processing does
not call for extremely high pressures for consolidation as are
Bicine (N,N-bis(2-hydroxyethyl) glycine) is already present
in this material system as a zirconium complexing agent used

December 2011
Hydrothermal Synthesis of Nano-Scale Zirconia and YTZP
4243
particles, which provides an electrostatic barrier to irrevers-
ible agglomeration.49
1YTZP Raman Spectrum
50000
45000
40000
35000
30000
25000
20000
15000
Deionized water was chosen as a control solution. All
washes were conducted at a pH of 9, which is a minimum
point in yttrium solubility in an aqueous environment.25–27
Washed samples were also dispersed in ethanol, adjusted to
pH 9, to observe the effect of compressive capillary forces on
aggregate strength during drying in a future study. Drying
from ethanol, with a lower surface tension than water, was
expected to decrease the compressive stress exerted on parti-
cles during solvent removal and, therefore, lower the average
aggregate strength.
Particle size distributions from dynamic light scattering
measurements given in Fig. 2 illustrate the effect of wash
solution additive on dispersion. Modes at 26, 32, and 41 nm
correspond to the measured size of primary particles of
YTZP washed and dispersed in water, bicine, and oxalate,
respectively. Modes at 20, 29, and 36 nm correspond to the
measured size of primary particles of YTZP washed in water,
oxalate, and bicine, respectively, and re-dispersed in ethanol.
Variations in mode position may be attributed to differences
in double layer thickness.34,35,50
Charged particles in suspension attract counterions of
opposite charge to balance the charge and maintain electro-
neutrality with the surrounding suspension. Counterions may
adsorb to particle surfaces and may also congregate around
the particle in a diffuse cloud. The former is called the Stern
layer while the latter is referred to as the Guoy–Chapman
layer. The double layer thickness can be calculated for
samples dispersed under different conditions with Eqs. (1)
and (2).
* - Tetragonal Zirconia
# - Unidentified (attributed to
residual organics)
*
*
*
*
*
#
*
800
600
400
200
0
Raman Shift (cm-1)
Fig. 5. Raman spectrum (using
a
785.493 nm laser excitation
source) of 1YTZP dry-pressed powder exhibits a pattern verifying
the presence of the tetragonal phase with peaks at 145, 268, 314,
463, 604 (shoulder), and 639 cmꢀ1 44 The peak at 418 cmꢀ1 is
attributed to residual bicine on particle surfaces.
.
Table I. Typical Impurity Concentrations (in ppm) Present
in 1YTZP Samples Washed in Ethanol (Zr-735-E) and Water
(Zr-735-W and Zr-741)
Zr-735-E
Zr-735-W
Zr-741
As each sample was diluted into pH-adjusted deionized
water or ethanol for the particle size measurement, the ionic
strength of each sample was around 2.5 9 10ꢀ4 mol/L (see
Table II). Double layer thickness calculated with Eq. (1) for
samples dispersed in water and ethanol was 19 and 11 nm,
respectively. Xu50 measured particle size shifts between 4 and
8 nm for similar double layer thicknesses. Particle size shift
values for YTZP samples are summarized in Table II.
In general, the double layer effects are greater for samples
containing additives, such as oxalate and bicine. This is not
unexpected as a diffuse layer of these ions surrounding a
particle represents electrostatic heterogeneities that interact
with the light used to interrogate the sample. Samples
washed without additives do not contain these heterogene-
ities, so the measured hydrodynamic radius gives a more
representative value of particle size.
Na2O
MgO
Al2O3
SiO2
P2O5
S
0.1
0.3
0.5
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
0.1
<0.1
0.2
0.3
<0.1
<0.1
0.1
<0.1
<0.1
<0.1
<0.1
0.2
<0.1
0.2
0.3
<0.1
0.1
0.1
<0.1
<0.1
<0.1
<0.1
0.2
<0.1
<0.1
<0.1
<100
<100
<100
<100
<100
<100
<200
<100
<100
<100
<100
<200
<300
Cl
K2O
CaO
TiO2
MnO
Fe2O3
BaO
V
Cr
Ni
Cu
Zn
As
Sn
Pb
Mo
Sr
U
Th
<0.1
<0.1
<0.1
<0.1
<100
<100
<100
<100
<100
<100
<200
<100
<100
<100
<100
<200
<300
<100
11 000
(0.77 mol%)
<100
<100
<100
<100
<100
<100
<200
<100
<100
<100
<100
<200
<300
<100
14 000
(0.99 mol%)
Following the recommendation of Xu,50 these observa-
tions illustrate the need to eliminate double layer effects,
either by compression with a high ionic strength medium or
by elimination of adsorbing ions, to obtain accurate particle
size measurements using dynamic light scattering.
Average agglomeration number, as defined earlier,
estimates the average number of primary particles in an
agglomerate.30 The packing fraction within an agglomerate
can be as low as 0.15 for a flocculated suspension, or as high
as 0.74 for close packed spherical particles.30 For random
close packing, a value of 0.6 is typically used. Hydrodynamic
Nb
Rb
Y
<100
14 000
(0.99 mol%)
diameter (DDLS) and equivalent spherical diameter (dBET
)
should be compared on a volume basis. Dynamic light scat-
tering data are presented as a volume distribution, and the
equivalent spherical diameter can be converted from
a
surface area basis to a volume basis (DBET) with Eq. (7).30
during synthesis.18 Evidence suggests that during and after
precipitation, bicine acts as a protective colloid preventing
irreversible agglomeration of newly formed particles.48 Also,
the branched, flexible structure of bicine may allow it to act
as a lubricant among particles during dry compaction.
Oxalic acid, another wash additive, has been shown to
be effective in suppressing the leaching of yttrium from
YTZP particles.49 It has also been shown that adsorbed
oxalate ions impart significant surface charge to YTZP
AAN values calculated with measured data are summa-
rized in Table II; if the effect of the double layer on the
measured hydrodynamic diameter is taken into account,
then actual AAN values for each wash solution are likely
to be close to 1. Therefore, good dispersion can be achieved
with any of the chosen wash solutions. It should be noted
that samples washed in water or bicine solution and dis-
persed in ethanol display broader particle size modes at lar-
ger sizes, which indicate that while primary particles are

4244
Journal of the American Ceramic Society—Szepesi et al.
Vol. 94, No. 12
Table II. Selected Characteristics of 1Y-ZrO2 Powder Synthesized by Hydrothermal Precipitation
Specific surface
area (m2/g)
DBET
,
Particle size shift,
Average agglomeration
number
Sample (YTZP)
Wash treatment
DDLS (nm)
dBET (nm)
Vol-basis (nm)
DDLS–DBET (nm)
Zr-781-B
Zr-781-BE
Bicine(aq) wash
Bicine(aq) wash,
29
33
108
124
9.2
8.0
17.9
15.6
11.1
17.4
2.5
5.7
Dispersed in ethanol
Oxalate(aq) wash
Oxalate(aq) wash,
dispersed in ethanol
DI water wash
Zr-781-O
Zr-781-OE
35
29
96
113
10.4
8.8
20.2
17.1
14.8
11.9
3.1
2.9
Zr-781-W
Zr-781-WE
23
18
124
132
8.0
7.6
15.6
14.7
7.4
3.3
1.9
1.1
DI water wash,
dispersed in ethanol
Table III. Solution Parameters Used for Calculations of Particle Interaction Energies
Deionized water, pH 9
Oxalic acid, pH 9
Bicine, pH 9
[H+] (mol/L)
10ꢀ9
10ꢀ5
10ꢀ9
10ꢀ5
0.17
0.01
–
0.42
0.21
ꢀ39
10ꢀ9
10ꢀ5
[OHꢀ] (mol/L)
[Tetramethyl ammonium+] (mol/L)
[Oxalate2ꢀ] (mol/L)
[Bicine+/ꢀ] (mol/L)
Conductivity (S/m)
Ionic strength (mol/L)
Zeta potential (mV)
4.9 9 10ꢀ4
–
2.5 9 10ꢀ2
–
0.01
–
1.4 9 10ꢀ3
2.5 9 10ꢀ4
ꢀ30
0.8
4.5 9 10ꢀ2
ꢀ37
dispersed in the suspension, some degree of agglomeration
is occurring.
Total Interaction Energy Summary,
10nm YTZP
The degree of dispersion was further evaluated through
measurements of zeta potential. Zeta potential determina-
tions permit the calculation of particle–particle interaction
energies using modified DLVO theory, which states that the
total energy of interaction between two charged particles is
the sum of the attractive (van der Waals) and repulsive (elec-
trostatic, steric) energies between them. The Stabil© computer
program24 calculates the total interaction energies and forces
based on several different models of interaction geometry. As
the particles in question can be approximated as nano-scale
spheres, the small ja (<0.5) spherical model was used.
Parameters necessary for the calculation, such as lattice
parameter51 and Hamaker constant,28 were taken from the
literature as 5.1 A and 8.8 9 10ꢀ20 J, respectively. Particle
radius was taken to be 5 nm, and the adsorbed layer thick-
ness was 0.37 or 0.75 nm for oxalate and bicine wash solu-
tions, respectively. Zeta potential was measured, and solution
ionic strength was calculated from conductivity data taken
during these measurements, which is given in Table III, and
equivalent conductivity values for each ionic species taken
from the literature.52 It should be noted that bicine is a zwit-
terion with both positive and negative charge at the working
pH value53 and, therefore, will not add to the overall con-
ductivity. It should be taken into account, however, in calcu-
lating solution ionic strength.
8
6
Water, pH 9
Oxalic Acid, pH 9
Bicine, pH 9
Oxalic Acid/Water, pH 9
4
2
0
5
0
10
20
30
40
-2
-4
-6
-8
0
0
2
4
6
8
10
-5
-10
-15
-20
Water, pH 9
Oxalic Acid, pH 9
Bicine, pH 9
Oxalic Acid/Water, pH 9
Separation Distance (nm)
Separation Distance (nm)
Fig. 6. The Stabil© computer program24 was used to calculate
interaction energy curves for YTZP washed in pH 9 deionized water,
pH 9 0.01M bicine solution, and pH 9 0.01M oxalic acid solution.
Repulsive and attractive energies are added to compute the total
interaction energy based on the spherical particle model. YTZP
washed with oxalate solution and dispersed in water displays the
greatest energy barrier to agglomeration due to the high zeta
potential and low ionic strength. The inset illustrates a steric barrier
to agglomeration when the energy curve ends at a distance of twice
the diameter of the adsorbed organic molecule (bicine or oxalate).
The interaction energy curves calculated for YTZP in the
three wash solutions are given in Fig. 6. For a suspension to
maintain good dispersion, a positive energy barrier greater
than thermal energy (approximately 3–4 kT) should exist
among suspended particles. In the YTZP–water system, this
energy barrier exists, indicating that particles will tend not to
agglomerate if they collide. For nano-scale particulates, even
steric barriers as small as those provided by the bicine
(0.75nm) and oxalate (0.35 nm) provide an electrosteric bar-
rier to agglomeration by truncating the short range van der
Waals interactions at a distance of twice the diameter of the
adsorbed molecule. In the YTZP–oxalate and YTZP–bicine
systems, there are no positive energy barriers, indicating a
net attraction between particles, but there is a finite energy
minimum provided by the oxalate and bicine adsorbed
layers, illustrated in the inset of Fig. 6. Thus, the size of the
adsorbed oxalate ions and bicine molecules on the 10 nm
diameter provides electrosteric dispersion even though no net
repulsive energy barrier is present.
The value of a reflects the likelihood that two particles will
collide with enough energy to overcome the repulsive energy
barrier and come to a separation distance where the attrac-
tive van der Waals interactions dominate the overall particle
interactions. The a value is calculated as the ratio of the
attractive energy (integrated over the separation distance
between two particles) to the total energy (also integrated
over the separation distance between two particles). As noted

December 2011
Hydrothermal Synthesis of Nano-Scale Zirconia and YTZP
4245
Effect of Wash Solution on Aggregation Probability
References
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1G. S. A. M. Theunissen, J. S. Bouma, A. J. A. Winnubst, and A. J. Burg-
graaf, “Mechanical Properties of Ultra-Fine Grained Zirconia Ceramics,”
J. Mater. Sci., 27 [15] 4429–38 (1992).
2F. F. Lange, “Transformation Toughening. Part 3. Experimental Observa-
tions in the ZrO2-Y2O3 System,” J. Mater. Sci., 17, 240–6 (1982).
3A. H. Chokshi, A. K. Mukherjee, and T. G. Langdon, “Superplasticity in
Advanced Materials,” Mater. Sci. Eng. R Rep., 10 [6] 237–74 (1993).
4K. Maca, M. Trunec, and P. Dobsak, “Bulk Zirconia Nanoceramics Pre-
pared by Cold Isostatic Pressing and Pressureless Sintering,” Rev. Adv. Mater.
Sci., 10 [1] 84–8 (2005).
5I. Santacruz, K. Anapoorani, and J. Binner, “Preparation of High Solids
Content Nanozirconia Suspensions,” J. Am. Ceram. Soc., 91 [2] 398–405 (2008).
6M. J. Mayo and M. Ciftcioglu, “Issues in the Processing of Bulk Nanocrys-
talline Ceramics for Structural Applications”; pp. 545–50 in Materials
Research Society Symposium Proceedings, Vol. 206, Clusters and Cluster-
Assembled Materials, Edited by R. S. Averback, J. Bernholc, and D. L. Nel-
son, Materials Research Society, Pittsburgh, PA, 1991.
7B. Xia, L. Duan, and Y. Xie, “ZrO2 Nanopowders Prepared by Low-Tem-
perature Vapor-Phase Hydrolysis,” J. Am. Ceram. Soc., 83 [5] 1077–80 (2000).
8V. V. Srdic, M. Winterer, and H. Hahn, “Sintering Behavior of Nanocrys-
talline Zirconia Doped With Alumina Prepared by Chemical Vapor Synthe-
sis,” J. Am. Ceram. Soc., 83 [8] 1853–60 (2000).
Water, pH 9
Oxalic Acid, pH 9
Bicine, pH 9
Oxalic
Acid/Water, pH 9
Wash Solution
Fig. 7. The Stabil© computer program24 was used to calculate the
probability of agglomeration based on the total interaction energy
and attractive interaction energy curves. The agglomeration
probability, a, is the likelihood that two particles will collide with
enough energy to come to a separation distance where the attractive
van der Waals interactions dominate. YTZP washed in water and
9W. Li and L. Gao, “Nano ZrO2 (Y2O3) Particles Processing by Heating of
Ethanol-Aqueous Salt Solutions,” Ceram. Int., 27, 543–6 (2001).
10J. Tartaj, E. Lachowski, J. F. Fernandez, C. Moure, and P. Duran,
“Microstructure Development and Densification of Ultrafine Y-TZP Mono-
liths Obtained by Seeding-Assisted Chemical Coprecipitation,” J. Euro. Ceram.
Soc., 20, 169–76 (2000).
11S.-F. Yin and B.-O. Xu, “On the Preparation of High-Surface-Area Nano-
Zirconia by Reflux-Digestion of Hydrous Zirconia Gel in Basic Solution,”
ChemPhysChem, 4 [3] 277–81 (2003).
dispersed in water or oxalic acid at pH
aggregation probability.
9 display the lowest
12S. Shukla, S. Seal, R. Vij, and S. Bandyopadhyay, “Effect of HPC and
Water Concentrtion on the Evolution of Size, Aggregation, and Crystallization
of Sol-Gel Nano Zirconia,” J. Nanopart. Res., 4, 553–9 (2002).
in Fig. 7, the aggregation probability is highest for YTZP
particles washed and dispersed in oxalic acid and bicine solu-
tions, 0.90 and 1.0, respectively; a is lowest for YTZP parti-
cles washed and dispersed in DI water or washed in DI
water and dispersed into oxalic acid solution, 0.09 and 0.06,
respectively.
13C. Laberty-Robert, F. Ansart, C. Deloget, M. Gaudon, and A. Rousset,
“Powder Synthesis of Nanocrystalline ZrO2-8%Y2O3 via
a Polymerization
Route,” Mater. Res. Bull., 36, 2083–101 (2001).
14C. Laberty-Robert, F. Ansart, C. Deloget, M. Gaudon, and A. Rousset,
“Dense Yttria Stabilized Zirconia: Sintering and Microstructure,” Ceram. Int.,
29, 151–8 (2003).
15T. Tsukada, S. Venigalla, A. A. Morrone, and J. H. Adair, “Low-Temper-
ature Hydrothermal Synthesis of Yttrium-Doped Zirconia Powder,” J. Am.
Ceram. Soc., 82 [5] 1169–74 (1999).
From these calculations, it is apparent that both the
YTZP–water and YTZP–bicine systems promote dispersion.
However, under these conditions, leaching of dopant is not
prevented.49 The major drawback of the YTZP–oxalate sys-
tem is its ionic strength, which is the greatest of the three
wash solutions. Particles in this system exhibit the greatest
zeta potential; however, a high ionic strength medium com-
presses the double layer surrounding the charged particles,
thus decreasing the repulsive energy between them. By
decreasing the ionic strength, surface charge compensation
by counterions extends further into the surroundings, thus
increasing the double layer thickness and the repulsive energy
barrier, as illustrated in Fig. 6. Therefore, to simultaneously
prevent dopant leaching and promote dispersion, YTZP par-
ticles were first washed in pH 9 oxalic acid and dispersed in
the final step into pH-adjusted deionized water.
16R. R. Piticescu, C. Monty, D. Taloi, A. Motoc, and S. Axinte, “Hydro-
thermal Synthesis of Zirconia Nanomaterials,” J. Euro. Ceram. Soc., 21,
2057–60 (2001).
17L. Zych and K. Haberko, “Zirconia Nanopowder – Its Shaping and Sin-
tering,” Solid State Phenom., 94, 157–64 (2003).
18R. A. Kimel and J. H. Adair, “Aqueous Synthesis at 200 Deg. C of Sub-
10 Nanometer Yttria Tetragonally Stabilized Zirconia Using a Metal-Ligand
Approach,” J. Am. Ceram. Soc., 88 [5] 1133–8 (2005).
19B. J. Intorre and A. E. Martel, “Zirconium Complexes in Aqueous Solu-
tion. I. Reaction with Multidentate Ligands,” J. Am. Chem. Soc., 82 [1] 358–
64 (1960).
20B. J. Intorre and A. E. Martel, “Aqueous Zirconium Complexes. II.
Mixed Chelates,” J. Am. Chem. Soc., 83 [9] 3618–23 (1961).
21B. J. Intorre and A. E. Martel, “Zirconium Complexes in Aqueous Solu-
tion. III. Estimation of Formation Constants,” Inorg. Chem.,
3 [1] 81–7
(1964).
22B. Derjaguin and L. Landau, “Theory of the Stability of Strongly Charged
Lyophobic Sols and of the Adhesion of Strongly Charged Particles in Solu-
tions of Electrolytes,” Prog. Surf. Sci., 43 [1–4] 30–59 (1993).
23E. J. W. Verwey and J. T. G. Overbeek, “Long Distance Forces Acting
Between Colloidal Particles,” Trans. Faraday Soc., 42B, 117–31 (1946).
24STABIL(c), University of Florida, Gainesville, FL, 1996.
IV. Conclusions
By modifying
a
hydrothermal technique, crystalline,
25J. H. Adair, R. P. Denkewicz, and F. J. Arriagada, “Precipitation and
In-Situ Transformation in the Hydrothermal Synthesis of Crystalline Zirco-
nium Dioxide”; pp. 135–45 in Ceramic Transactions, Vol. 1, Ceramic Powder
Science II, Edited by G. L. Messing, E. R. Fuller, and H. Hausner. The Amer-
ican Ceramic Society, Westerville, OH, 1988.
8–10 nm, tetragonal 1YTZP has been produced with a low
impurity content and a yield as high as 100 g/L. The crystal
structure was established with XRD and Raman spectros-
copy, and the particle size was confirmed using dynamic light
scattering, BET surface area measurements, TEM observa-
tions, and XRD peak broadening. The enhanced tendency
for agglomeration in high-yield batches was successfully sup-
pressed by the bicine protective colloid. An effective washing
and dispersion procedure was identified in which precipitated
suspensions were washed with aqueous oxalic acid, pH
adjusted to a value of 9, and dispersed in the final step into
pH-adjusted deionized water. The effect of double layer
thickness on measurement of hydrodynamic radius using
dynamic light scattering was discussed. Taking into account
double layer effects on the measured particle size, AAN val-
ues are near unity, indicating good dispersion with this pro-
cedure. Interaction energy curves based on DLVO theory
were calculated to illustrate the observed dispersion behavior.
26J. H. Adair, H. G. Krarup, S. Venigalla, and T. Tsukada, “A Review of
the Aqueous Chemistry of the Zirconium - Water System to 200 Degrees C”;
pp. 101–12 in Aqueous Chemistry and Geochemistry of Oxides, Oxyhydroxides,
and Related Materials, Materials Research Society Proceedings, Vol. 432, Edi-
ted by J. A. Voight, T. E. Wood, B. C. Bunker, W. H. Casey, and L. J. Cros-
sey. Materials Research Society, Pittsburgh, PA, 1997.
27T. Tsukada, S. Venigalla, and J. H. Adair, “Crystallization of Zr Hydrox-
ide to ZrO2 at/or Below 100 Degrees C”; pp. 123–39 in Ceramic Transactions,
Vol. 54, Science, Technology, and Applications of Colloidal Suspensions, Edited
by J. H. Adair, J. A. Casey, C. A. Randall, and S. Venigalla. The American
Ceramic Society, Westerville, OH, 1995.
28L. Bergstrom, A. Meurk, H. Arwin, and D. J. Rowcliffe, “Estimation of
Hamaker Constants of Ceramic Materials From Optical Data Using Lifshitz
Theory,” J. Am. Ceram. Soc., 79 [2] 339–48 (1996).
29Encyclopedia of Materials Characterization: Surfaces, Interfaces, Thin Films.
Edited by C. R. Brundle, J. C. A. Evans, and S. Wilson. Butterworth-Heine-
mann, Stoneham, MA and Manning Publishing Co., Greenwich, CT, 1992

4246
Journal of the American Ceramic Society—Szepesi et al.
Vol. 94, No. 12
30N. E. Antolino, G. Hayes, R. Kirkpatrick, C. L. Muhlstein, M. I. Frecker,
41Y. Liang, N. Gregory, and J. G. P. Binner, “Green Forming With Nano-
particles,” Key Eng. Mater., 264–268, 2319–22 (2004).
E. M. Mockensturm, and J. H. Adair, “Lost Mold Rapid Infiltration Forming
of Mesoscale Ceramics: Part 1, Fabrication,” J. Am. Ceram. Soc., 92 [S1] S63–
9 (2009).
42W. H. Rhodes, “Agglomerate and Particle Size Effects on Sintering Yttria-
Stabilized Zirconia,” J. Am. Ceram. Soc., 64 [1] 19–22 (1981).
43O. L. Khasanov, O. V. Karban, and E. S. Dvilis, “Investigation of Struc-
tural Hierarchy of Nanoceramics Compacted by Dry Pressing Under Powerful
Ultrasound Action,” Key Eng. Mater., 264–268, 2327–30 (2004).
44W. Li and L. Gao, “Compacting and Sintering Behavior of Nano ZrO2
Powders,” Scr. Mater., 44, 2269–72 (2001).
31R. P. Denkewicz Jr., K. S. TenHuisen, and J. H. Adair, “Hydrothermal
Crystallization Kinetics of m-ZrO2 and t-ZrO2,” J. Mater. Res., 5 [11] 2698–
705 (1990).
32G. A. Schumacher and T. G. M. VanDeVen, “Brownian Motion of
Charged Colloidal Particles Surrounded by Electric Double Layers,” Faraday
Discuss. Chem. Soc., 83, 75–85 (1987).
45C.-Y. Xie, L.-D. Zhang, Z.-G. Zhu, Z.-Y. Chen, and Y. Li, “Low Tem-
perature Internal Friction Spectrum of Nano ZrO2 Solids,” Sci. China, Ser. A,
35 [11] 1336–42 (1992).
33R. Pecora, “Dynamic Light Scattering Measurement of Nanometer Parti-
cles in Liquids,” J. Nanopart. Res., 2, 123–31 (2000).
34J. L. Viota, J. deVicente, M. M. Ramos-Tejada, and J. D. G. Duran,
“Electrical Double Layer and Rheological Properties of Yttria-Stabilized
Zirconia Suspensions in Solutions of High Molecular Weight Polyacrylic Acid
Polymers,” Rheol. Acta, 43, 645–56 (2004).
46C. D. Sagel-Ransijn, A. J. A. Winnubst, B. Kerkwijk, A. J. Burggraaf,
and H. Verweij, “Production of Defect-Poor Nanostructured Ceramics of
Yttria-Zirconia,” J. Euro. Ceram. Soc., 17, 831–41 (1997).
47C. Xie, L. Zhang, C. Mo, Z. Zhu, Z. Chen, Y. Qian, and Y. Li, “Internal
Friction and Modulus of Nano-ZrO2 Solids,” J. Appl. Phys., 72 [8] 3447–50
(1992).
35R. G. Iacocca and R. M. German, “A Comparison of Powder Particle
Size Measuring Instruments,” Int. J. Powder Metall., 33 [8] 35–48 (1997).
36E. Fernandez-Lopez, V. Sanchez-Escribano, M. Panizza, M. M. Carnasci-
ali, and G. Busca, “Vibrational and Electronic Spectroscopic Properties of
Zirconia Powders,” J. Mater. Chem., 11, 1891–7 (2001).
48R. A. Kimel, “Aqueous Synthesis and Processing of Nanosized Yttria
Tetragonally Stabilized Zirconia”; Ph.D. thesis in Materials Science and Engi-
neering, The Pennsylvania State University, University Park, PA, 2002.
49R. A. Kimel and J. H. Adair, “Aqueous Degradation and Chemical Pas-
sivation of Yttria-Tetragonally Stabilized Zirconia at 25°C.,” J. Am. Ceram.
Soc., 85 [6] 1403–8 (2002).
37D. J. Green, An Introduction to the Mechanical Properties of Ceramics,
Cambridge University Press, Cambridge, UK, 1998.
38M. J. Mayo, “Nanocrystalline Ceramics for Structural Applications:
Processing and Properties”; pp. 361–85 in Nanostructured Materials: Science
and Technology, Edited by G.-M. Chow and N. I. Noskova. Kluwer Academic
Publishers, Boston, MA, 1998.
50R. Xu, “Shear Plane and Hydrodynamic Diameter of Microspheres in Sus-
pension,” Langmuir, 14, 2593–7 (1998).
51J.-W. Jang, H.-K. Kim, and D.-Y. Lee, “The Effect of Tetravalent
Dopants on the Unit Cell Volume of 2Y-TZP and 8Y-SZ,” Mater. Lett., 58,
1160–3 (2004).
39A. O. Boschi and E. Gilbart, “Wet Forming Processes as a Potential Solu-
tion to Agglomeration Problems”; pp. 73–93 in Advanced Ceramic Processing
and Technology 1, Edited by J. G. P. Binner. Noyesd Publications, Park Ridge,
NJ, 1990.
52CRC Handbook of Chemistry and Physics, 81st edition. Edited by D. R.
Lide. CRC Press, Boca Raton, FL, 2000
40H. Nienburg and F. Harbach, “Pressure Filtration of Fine Ceramic Sus-
pensions”; pp. 321–7 in Ceramic Powder Science IV, Edited by S. Hirano,
G. L. Messing, and H. Hausner. American Ceramic Society, Westerville, OH,
1991.
53H. A. Azab, Z. A. Khafagy, A. Hassan, and A. M. El-Nady, “Medium
Effect on the Second-Stage Dissociation Constant of N,N-Bis(2-Hydroxyethyl)
Glycine,” J. Chem. Eng. Data, 39, 599–601 (1994).
h