Mechanical and magnetic properties of novel yttria-stabilized tetragonal zirconia/Ni nanocomposite prepared by the modified internal reduction method

Article abstract of DOI:10.1111/j.1551-2916.2005.00243.x

Dense ceramic/metal nanocomposite has been fabricated by internal reduction method, which includes a two-step process: sintering of ceramic-metal oxide solid solution and subsequent heat treatment in a reducing atmosphere to precipitate metal nanoparticles. This novel technique has been applied to yttriastabilized tetragonal zirconia (Y-TZP) and nickel oxide (NiO) system to fabricate Y-TZP/Ni nanocomposite. Dense Y-TZP and 0.3 mol% NiO solid solution ceramic was successively prepared by the pressureless sintering, and Y-TZP/Ni was fabricated by the internal reduction treatment. The obtained Y-TZP/Ni nanocomposite possessed characteristic intragranular nano-structure with nano-sized metallic Ni particles of around 20 nm. Fracture toughness of both the solid solution and nanocomposite was remarkably improved because of the solid solution of NiO into Y-TZP and resultant destabilization of the tetragonal phase, and the Y-TZP/Ni nanocomposite was still destabilized by the remaining nickel solution after the reduction. The nano-composite exhibited ferromagnetism, while the Y-TZP-NiO solid solution had diamagnetic nature. Comparison of saturation magnetization values revealed that 39.5 at.% of introduced nickel was reduced to metallic nanoparticle, proving the existence of residual NiO solute in zirconia that contributed to higher toughness value than the monolithic Y-TZP. It is concluded that the introduced internal reduction method is a suitable process to achieve multifunctional ZrO₂/Ni nanocomposite with high toughness and coexistent magnetic characteristic.

Full text of DOI:10.1111/j.1551-2916.2005.00243.x

J. Am. Ceram. Soc., 88 [6] 1468–1473 (2005)
DOI: 10.1111/j.1551-2916.2005.00243.x
ournal
J
Mechanical and Magnetic Properties of Novel Yttria-Stabilized
Tetragonal Zirconia/Ni Nanocomposite Prepared by the Modified
Internal Reduction Method
Hiroki Kondo, Tohru Sekino,*,w Norihito Tanaka, Tadachika Nakayama, Takafumi Kusunose,* and
Koichi Niihara**
The Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka 567-0047, Japan
Dense ceramic/metal nanocomposite has been fabricated by
internal reduction method, which includes a two-step process:
sintering of ceramic–metal oxide solid solution and subsequent
heat treatment in a reducing atmosphere to precipitate metal
nanoparticles. This novel technique has been applied to yttria-
stabilized tetragonal zirconia (Y-TZP) and nickel oxide (NiO)
system to fabricate Y-TZP/Ni nanocomposite. Dense Y-TZP
and 0.3 mol% NiO solid solution ceramic was successively pre-
pared by the pressureless sintering, and Y-TZP/Ni was fabri-
cated by the internal reduction treatment. The obtained Y-TZP/
Ni nanocomposite possessed characteristic intragranular nano-
structure with nano-sized metallic Ni particles of around 20 nm.
Fracture toughness of both the solid solution and nanocomposite
was remarkably improved because of the solid solution of NiO
into Y-TZP and resultant destabilization of the tetragonal
phase, and the Y-TZP/Ni nanocomposite was still destabilized
by the remaining nickel solution after the reduction. The nano-
composite exhibited ferromagnetism, while the Y-TZP–NiO
solid solution had diamagnetic nature. Comparison of satura-
tion magnetization values revealed that 39.5 at.% of introduced
nickel was reduced to metallic nanoparticle, proving the exist-
ence of residual NiO solute in zirconia that contributed to higher
toughness value than the monolithic Y-TZP. It is concluded that
the introduced internal reduction method is a suitable process
to achieve multifunctional ZrO2/Ni nanocomposite with high
toughness and coexistent magnetic characteristic.
new functions combined with required mechanical proper-
ties.4,10,11 A fracture strength of 1090 MPa was reported for
the Al2O3/Ni nanocomposite with 5 vol% Ni fabricated by re-
duction and hot-pressing method, which exceeds more than
twice the fracture strength of monolithic Al2O3 prepared by
the same procedure.4 Furthermore, the introduction of Ni nano-
particles into the sintered ceramic resulted in addition of ferro-
magnetic behavior4 and accompanied magnetic remote sensing
capability of stress and/or fracture10 of the composite. Given the
above-described success, the Ni nanoparticles were dispersed
within yttria-stabilized tetragonal zirconia (Y-TZP), which led
to remarkable improvement of fracture strength (1.9 GPa)
achieved for Y-TZP/Ni nanocomposite system with 1–2 vol%
of Ni.12 Thus, the combination of enhanced strength and
ferromagnetic properties demonstrated by the obtained new
nanostructured material enabled to elaborate on its multifunc-
tionality—the possibility of its application for stress sensor (see
Kondo et al.12).
The ceramic/metal nanocomposites are often prepared using
the reduction and hot-pressing method4,8–16 in which the powder
mixtures of ceramic matrix and metal oxide are heated in hy-
drogen atmosphere to reduce metal oxide to metal, and subse-
quently hot pressed, as illustrated in Fig. 1(a). The advantage of
this method is the possibility of obtaining relatively fine disper-
sion of metal particles, in contrast to other processes that use
metal powder as the initial material.16
Furthermore, there were attempts to obtain ceramic/metal
nanocomposite powders by solution chemical processing,4,17–20
which provided ceramic/metal nanocomposites with better
nanostructure. Sekino et al.4 proposed a particular case of
such a method, where Ni(NO3)2 nitride was chosen as a source
material of metal dispersions. One should be aware, however,
that in the conventional reduction and sintering process the me-
tallic Ni tends to migrate and coalesce with other Ni particles
during sintering, since in the majority of ceramic/metal systems
the sintering temperature of ceramics (e.g., 14501 and 15001C for
Al2O3 and Y-TZP, respectively) is close to the melting point of
nickel (14551C).
In order to surpass the indicated difficulties, the present paper
aims to introduce the internal reduction method that might al-
low producing bulk Y-TZP/Ni nanocomposite with the finest Ni
particles (see Fig. 1(b)). The internal reduction method has al-
ready been applied to various mono- and mixed-oxide systems
such as Fe2O3,21 (A,B)O,22–24 transition metal (Cr, Ni, Fe)-
doped Al2O3,20,25–27 Fe-doped silicate,28 and so on, in order to
investigate reduction dynamics/kinetics and mechanisms, crys-
tallographic characteristics of metal precipitates within oxide
matrices, microstructures, and/or physical/mechanical proper-
ties. This method thus realizes ceramic/metal, or to be more
precise, oxide/metal composite microstructures. In the case of
Ni–Al2O3 composite,20,27 NiAl2O4 spinel compound was pre-
pared first and then partially reduced to form the composite.
However, size distribution of precipitated metal particles in the
alumina matrix was large; some precipitates were in nanometer
I. Introduction
REMARKABLE improvement of mechanical properties was
achieved in the ceramic-based nanocomposite system by
A
dispersing nano-scaled second-phase particles into ceramic ma-
trix.1,2 Such a dispersed second phase in a particular nanocom-
posite structure enables to drastically alter the behavior of
material that exhibits new functions3–6 such as super-plasticity
or high machinability like metal. The fabrication of the ceramic/
metal nanocomposite system is nowadays performed in such a
way as to achieve a proper balance between enhanced mechan-
ical properties and addition of new functions that are frequently
based on electrical and magnetic behavior.7–9 The Al2O3/Ni ce-
ramic/metal nanocomposite system that has been widely studied
already is a good example of the system developed toward
V. Jayaram—contributing editor
Manuscript No. 10489. Received August 29, 2003; approved November 3, 2004.
This work was supported by the Ministry of Education, Culture, Sports, Science and
Technology, Japan, under a Grant-in-Aid for JSPS Research Fellowships for Young Sci-
entists (H. Kondo) No. 1907, and partly supported by the Ministry of Economy, Trade and
Industry, Japan, as part of the Synergy Ceramics Project.
*Member, American Ceramic Society.
**Fellow, American Ceramic Society.
wAuthor to whom correspondence should be addressed. e-mail: sekino@sanken.osaka-
u.ac.jp
1468

June 2005
Mechanical and Magnetic Properties of Y-TZP/Ni Nanocomposite
1469
cern whether the selected Ni content would be too low. How-
ever, from the previous experience with nano-sized second
phase, we suspected that even a small amount of nanoparticles,
less than 2%, may result in drastic improvement of properties.12
(a) Conventional method
(i.e, reduction and hot pressing)
Nanocomposite
(Coarse structure)
Metal dispersion
Ceramic matrix
(2) Powders and Specimens Preparation
Ceramic/metal oxide
mixed powder
The initial mixture used for fabrication of the solid solution and
consequent nanocomposite contained tetragonal zirconia pow-
der stabilized with 3 mol% yttria (3Y-TZP, 0.2 mm, Sumitomo
Osaka Cement Co. Ltd., Tokyo, Japan) and NiO powder (350
mesh, Nakalai Tesque Inc., Kyoto, Japan). These powders were
wet-ball milled using ethanol and zirconia ball as a solvent and
mixing media, respectively. Dried mixtures were then dry-ball
milled again. The NiO content was chosen to be 0 and 0.3 mol%
to Y-TZP, which corresponded to 0 and 0.1 vol% Ni in the final
products, respectively. The mixed powder was uniaxially pressed
under the pressure of 20 MPa into a pellet (size of +15 mm ꢀ
4 mm) and subsequently cold isostatic pressed at 200 MPa. The
pellets were consolidated by pressureless sintering at 15001C in
air. In order to obtain a dense and perfect solid solution of the
Y-TZP and NiO, the sintering was carried out for 24 h (for de-
tails refer to Kondo et al.30). The specimen size of +12 mm ꢀ
3 mm was then obtained after the sintering.
Pressureless
sintering
Reduction
Solid solution
Nanocomposite
(Fine structure)
(b) Internal reduction
Fig. 1. Schematic drawing of materials preparation concept for the
yttria-stabilized tetragonal zirconia/Ni nanocomposite by the conven-
tional method (a) and the present internal reduction method (b).
scale and the others were large, micrometer sized.20 The mod-
ified internal reduction method presented in this research con-
sists of utilizing not only thermodynamic stability of doped
element but also high oxygen ionic mobility of zirconia ceramic,
which allows homogeneous reduction of selected element doped
into dense sintered body by forming nanometer-sized metal pre-
cipitates, but maintaining the characteristics of matrices such as
high strength and toughness.
Apart from the internal reduction method, the paper focuses
on the properties of the obtained materials. Hence, the micro-
structure developments and mechanical and magnetic properties
of Y-TZP/Ni nanocomposite fabricated by internal reduction
method were investigated, which indicates usefulness of this
fabrication technique. In addition, the present paper also dis-
cusses the effects of nickel oxide (NiO) addition on the phase
stability and mechanical properties of obtained materials, Y-
TZP–NiO solid solution, and internal-reduced Y-TZP/Ni nano-
composite.
For the internal reduction treatment, the sintered solid solu-
tion was heated for 2.5 h in a pure hydrogen (99.99%) atmos-
phere at 13001C. The temperature was chosen to secure fine
Ni precipitation and to avoid monolithic to tetragonal-phase
transformation of zirconia matrix (see the respective phase
diagram31) as well as grain growth of ZrO2 matrix.
(3) Material Characterization
The phase composition of the prepared nanocomposite was ex-
amined by X-ray diffraction (XRD, RAD-C system, Rigaku
Co., Tokyo, Japan) technique with CuKa radiation. The lattice
parameters were determined using high-purity silicon as the in-
ternal standard. The density of the composite was determined by
Archimedes’ immersion method using toluene. The characteri-
zation of mechanical properties focused on fracture toughness
that was determined by the indentation fracture (IF) method
using Vickers indentation under the maximum load of 196 N
and a loading duration of 15 s, while making use of an empirical
equation for Palmqvist cracks32 and also by the single-edge-
precracked-beam (SEPB) method.33 For the toughness determi-
nation, at least six indentations and five bending tests were per-
formed and averaged for the IF and SEPB methods, respectively.
The selected specimens were thermally etched for 10 min at
13001C in vacuum to facilitate the microstructural observations
performed by scanning electron microscopy (S-5000, Hitachi
Ltd., Tokyo, Japan), which enabled to determine average grain
size of Y-TZP matrix by image analysis software (Scion Image,
Scion Corporation, Frederick, MD).
In order to resolve the local phase composition in detail,
microprobe laser Raman spectroscopy (SPEX1482D, SPEX
Industries Inc., Edison, NJ, argon ion laser, l 5 514.5 nm, the
beam diameter of approximately 3 mm, power 100 mW) was
performed, since this technique proved to be well suited for
ZrO2-based systems.34 Interestingly, the combination of inden-
tation, Raman, and XRD techniques allowed to evaluate the
volume fraction of monoclinic phase around the indentation-
induced crack. This was accomplished using relative intensities
of the most distinctive Raman bands of zirconia: 148, 181, 192,
and 264 cmꢁ1, according to the formula:34
II. Experimental Procedure
(1) Modified Internal Reduction Method
The general outline of the present fabrication method consists of
two stages (see Fig. 1(b)). Firstly, the mixed powders of Y-TZP
and NiO are sintered to obtain the dense Y-TZP–NiO solid so-
lution. Secondly, the bulk Y-TZP–NiO solid solution is heated
in reductive atmosphere (presumably pure hydrogen) to reduce
the Ni21 ions and form Ni particles within the dense ceramic
matrix. The latter is realized mainly because of the oxygen ion-
transfer in zirconia polycrystal. The reductive atmosphere in-
volves the decrease in chemical potential of oxygen in the solid,
and results in the reduction of dissolved Ni21 ions in the Y-TZP.
Interestingly, the reduction of Ni21 ions appears to be internal,
in contrast to previously elaborated ZrO2 in Y-TZP.12,29 Indeed,
it is considerably easier to reduce Ni than ZrO2 despite the fact
that the ionic conductivity of the latter exceeds oxygen ionic
conductivity of Y-TZP.12 It is worth emphasizing that in the
case of solid solution, one deals with nickel ions instead of par-
ticles, which provides an opportunity to produce Ni particles
with a very small size.
The initial step in the present research was the investigation of
the Y-TZP–NiO solid solution prepared by pressureless sin-
tering (the detailed discussion of the results is given elsewhere30),
since it is believed that the properties of Y-TZP–NiO solid so-
lution strongly, affect the final product of internally reduced Y-
TZP/Ni nanocomposite.
Y-TZP–0.3 mol% NiO solid solution was successfully ob-
tained by pressureless sintering. The addition of a larger amount
of NiO resulted in microcracking of the sintered body because of
over-destabilization. As a consequence, the monolithic Y-TZP
and Y-TZP–0.3 mol% NiO solid solution that in certain aspects
correspond to Y-TZP/0.1 vol% Ni after complete reduction
were selected for the present investigation. It was of great con-
Im181 þ Im192
Cm
¼
;
(1)
FðIt148 þ It264Þ þ Im181 þ Im192
where Cm is the monolithic concentration, Im and It refer to the
integrated intensities of monoclinic and tetragonal bands, while
F stands for a factor required to convert the Raman intensity to

1470
Journal of the American Ceramic Society—Kondo et al.
Vol. 88, No. 6
the XRD intensity of the reference material; for the present
investigation, F 5 0.97 was used according to the previous
report.34
Free surface
Vickers indentation
Further, structural investigation was performed by transition
electron microscopy (TEM, H-8100, Hitachi Ltd.) with an
energy-dispersive X-ray spectrometer (PV9800 with ultra-thin
window detector, EDAX Inc., Mahwah, NJ) for the elemental
analysis of nanocomposites. Finally, the magnetic properties
of the Y-TZP/Ni nanocomposite were determined by the
super-conducting quantum interference magnetometer (MPMS2,
Quantum Design Inc., San Diego, CA) at room temperature,
which provided the amount of precipitated Ni particles by
comparing theoretical and measured saturation magnetization
values of the composites.
Palmqvist crack
50 µm
50 µm
Probe point A
Probe point B
Fig. 2. Schematic diagram illustrating the cracks introduced by Vickers
indentation and the positions of microprobe laser Raman spectroscopy
analysis.
III. Results and Discussion
(1) Phase Development
search. Moreover, the monolithic Y-TZP heat treated in the
same reduction atmosphere was prepared and tested for the sake
of comparison (refer to Table I that also contains the data for
the Y-TZP–NiO described elsewhere30).
The Y-TZP with 0.3 mol% NiO solid solution was densified
after pressureless sintering without any crack.30 After the reduc-
tion treatment of the sintered body at 13001C, the light brown
color of the Y-TZP–0.3 mol% NiO solid solution turned black,
while the monolithic Y-TZP changed from white to light gray.
The color change of the Y-TZP–0.3 mol% NiO system implies
reduction of Ni21 and precipitation of metallic nickel in the
sintered body. The color change of the monolithic Y-TZP to
light gray may be because of the increase in oxygen defects and/
or solid solution of carbon or hydrogen induced during reduc-
tion. Unfortunately, the XRD measurement for confirming the
existence of metallic Ni embedded in the matrix could not detect
the tiny amount of Ni phase or change of the lattice parameter.
In addition, no signal attributing to elemental Ni was detected
by preliminary X-ray photoelectron spectroscopy (XPS) inves-
tigation of the solid solution as well as the reduced sample, be-
cause of lesser quantity of the doped NiO than the detection
limit.
It is worth noticing that Table I contains the data of the av-
erage grain size of the zirconia matrix, since it is well known that
this parameter affects the toughness of tetragonal zirconia. One
may easily recognize that fracture toughness of the Y-TZP–0.3
mol% NiO solid solution measured by the SEPB method is
again higher than that of the Y-TZP monolith, which is attrib-
uted to the larger average grain size in solid solution and desta-
bilization of ZrO2 tetragonal phase by NiO doping.30 High
toughness of the Y-TZP/Ni nanocomposite, close to the values
obtained by both indentation and SEPB tests for the Y-TZP–0.3
mol% NiO solid solution, was unchanged after reduction. This
suggests that the complete reduction of NiO to metallic Ni did
not occur for the entire volume of NiO.
(2) Fracture Toughness
(3) Phase Transformation Behavior
In contrast to the XRD and XPS results, however, the fracture
toughness of Y-TZP measured by the indentation method (see
Table I) was remarkably influenced by the solid solution of only
0.3 mol% NiO to Y-YZP. Furthermore, a high value was main-
tained after the internal reduction treatment as listed in Table I.
The present authors were aware that the indentation method
frequently gave overestimated values of fracture toughness in
the case of zirconia ceramics,35 because of unusual R-curve be-
havior associated with transformation toughening of tetragonal
zirconia.36 As a consequence, the alternative SEPB method was
also employed for fracture toughness measurement in this re-
In order to determine the amount of transformed monoclinic
phase, microprobe laser Raman spectroscopy analysis was car-
ried out in a selected location around cracks induced by Vickers
indentation, as it is illustrated in Fig. 2. The three analyzed
points comprising the polished and, hence, stress-free surface,
point A being 50 mm distant from the crack root and point B
located 50 mm from the crack tip, were selected (Fig. 2), and the
results are listed in Table II. A very limited amount of mono-
clinic phase was detected at the location A of the monolithic Y-
TZP, since the tetragonal-to-monoclinic (t–m) transformation
zone was relatively narrow30 (note the analyzed area was located
50 mm away from the crack). In contrast, the Y-TZP–0.3 mol%
NiO solid solution exhibited higher amount of monoclinic phase
both in A and B positions than the monolithic Y-TZP. This can
be readily explained by the fact that solid solution of NiO de-
stabilizes tetragonal phase of the Y-TZP, which in turn tends to
transform to monoclinic phase, and as a result, the t–m zone
extends.30
Table I. Fracture Toughness and Average Grain Size of
Y-TZP Matrix for the Sintered Y-TZP Monolith, Y-TZP–
NiO Solid Solution and Y-TZP/Ni Nanocomposite Fabricated
by the Internal Reduction Method
Fracture toughness (MPa ꢂ m1/2
)
Average grain
size (nm)
Table II. Amount of Monoclinic Zirconia Phase on the Free
Surface and Near Cracks Calculated from Raman Spectroscopy
Data
Sample
IF method
SEPB method
Y-TZP monolith
Reduced Y-TZP
monolith
Y-TZP–0.3 mol%
NiO solid solution 9.9770.20
Y-TZP/Ni
6.3470.13
5.0570.14
4.9070.32
6.6070.12
6.7370.24
830
850
Amount of monoclinic phase
(vol%)
6.5370.14
1010
Sample
Free surface Point A Point B
Y-TZP monolith
Y-TZP–0.3 mol% NiO solid solution 0.0
Y-TZP/Ni nanocomposite 0.7
0.0
2.3 0.0
14.7 1.2
14.6 6.4
nanocomposite
9.4370.41
950
Y-TZP monolith was reduced to confirm the effect of hydrogen treatment
for comparison. Errors denote standard deviation of the measurement. Y-TZP,
yttria-stabilized tetragonal zirconia; IF, indentation fracture; SEPB, single-edge-
precracked-beam.
See Fig. 2 for the geometry of analyzed points. Y-TZP, yttria-stabilized te-
tragonal zirconia.

June 2005
Mechanical and Magnetic Properties of Y-TZP/Ni Nanocomposite
1471
A similar amount of monoclinic phase was detected in the Y-
TZP–0.3 mol% NiO solid solution and the Y-TZP/0.1 vol% Ni
nanocomposite that was a reduced sample of the former solid
solution (compare Tables I and II). The Raman result agrees
well with the fracture toughness measurements. This prompts us
to conclude that t–m transformation of NiO-doped tetragonal
phase is sensitive to stress and it is why the t–m transformation
zone was extended in the Y-TZP/Ni nanocomposite. Indeed, the
Raman analysis proved that a certain amount of NiO was still
present in the Y-TZP/0.1 vol% Ni nanocomposite even after
the internal reduction occurred, which suggested that the exact
amount of precipitated Ni would be lower than 0.1 vol%. De-
tails of this phenomenon will be discussed in the latter part.
(4) Nanostructure of Internal Reduced Specimen
The TEM micrographs of the nanocomposite, which was an
internally reduced material of the sintered solid solution, shown
in Fig. 3, reveal a characteristic structure with very fine particles
with sizes approximately 20 nm, mainly within the ZrO2 grains
and some at the grain boundary. As shown in Figs. 3(a) and (b),
there are two types of nanodispersion: one is a black-contrasted
particle (Fig. 3(a)) and the other is a black particle accompany-
ing the bright-contrasted region (Fig. 3(b)). These particles
contain nickel, as proved by EDX analysis. The elemental com-
position of the area with a bright contrast visible in Fig. 3(b) was
exactly the same as that of ZrO2 grains. Hence, the EDX results
prompted us to conclude that the contrast comes from a pore or
a part of ZrO2 that had many vacancies or defects. The high-
resolution image for the latter type of nanodispersion is shown
in Fig. 3(c). Since the shape of the discussed region was sur-
rounded by flat planes, it is believed that the contrast is related
to the precipitation of metallic Ni particles. When one assumes
that the absorbed contrast is caused by a pore, such a void may
be readily formed during reduction, since it involves volume de-
crease. Unfortunately, at the present stage of research, both ex-
planations of the origin of the contrast (pore and precipitation)
seem equally plausible.
The clear lattice image of the Ni nanoparticle was not con-
firmed in the high-resolution image (Fig. 3(c)). Nevertheless,
another interesting feature found from the high-resolution im-
age was that the facets consisted of low Miller index planes of
zirconia such as {101} and {110}. A similar feature has already
been reported by many researchers. In the case of Cr, metal
precipitation from Cr-doped Al2O3, facet planes consisting
mainly of a low-energy plane and crystallographic orientation
between intragranular Cr precipitates and parent Al2O3 crystal
were reported.25 The formation of this kind of facet and corre-
sponding shape can be related to the Wullf shape often observed
for an internal pore and/or a precipitate in solid and/or at wet
interface.37,38 It is therefore reasonable to think that formation
of flat hetero-interface between nano-sized intragranular Ni
metal and zirconia crystal as well as between pore and zircon-
ia might be governed mainly by the formation of the Wulff
structure during internal reduction, which minimized interfacial
energy between zirconia and Ni and the accompanying pore. In
addition, Cr precipitation at the grain boundary of Al2O3 was
reported not to have clear orientation.25 It is considered that the
lack of crystallographic orientation between internally reduced
Ni precipitates might be energetically disadvantageous in the
case of such a precipitation phenomenon.
Fig. 3. Transition electron microscopy micrographs of nano-sized Ni
particles precipitated within a yttria-stabilized tetragonal zirconia (Y-
TZP) grain in Y-TZP/Ni nanocomposite (a, b) and high-resolution im-
age of a Ni precipitate and surrounding zirconia crystal (c) prepared by
the internal reduction technique. Black and white arrows indicate Ni
particles and the bright-contrasted part, respectively.
(5) Magnetic Properties and the Amount of Reduced Ni
The magnetic properties of the materials prepared by the present
study exhibited an expected sequence. Hence, as is presented
in Fig. 4, the Y-TZP–0.3 mol% NiO solid solution possessed a
typical diamagnetic property that was of the same nature as
monolithic Y-TZP (not shown), despite the fact that the cubic
form of NiO had antiferromagnetism as its intrinsic magnetic
characteristic. Such a behavior exhibits neither precipitation of
NiO nor anti-ferromagnetic coupling between solute Ni21 ions
in tetragonal ZrO2 cell because of very low concentration of Ni
in Y-TZP (note that 0.3 mol% NiO was solid solute in Y-TZP).
In contrast, the Y-TZP/Ni exhibits ferromagnetic nature, as is
seen in Fig. 4. The fact that the ferromagnetic properties of the
nanocomposite are not shadowed by diamagnetism of the Y-

1472
Journal of the American Ceramic Society—Kondo et al.
Vol. 88, No. 6
In the case of the present study, Ni particle size (20 nm) was
smaller than the previous report (Ni with approximately 100 nm
within Y-TZP/Ni nanocomposites prepared by a reduction and
followed by sintering of TZP/NiO mixture),12 i.e., the surface
area of Ni dispersion was large, which inferred lower saturation
magnetization of the Y-TZP/Ni nanocomposite. One should
also notice that metallic Ni particles embedded in Y-TZP matrix
are next to oxygen ions of ZrO2 (the surface of Ni particle is
surrounded by an oxide layer), which contribute to the observed
reduction of Ms-value as is similar to the case of nanocrystalline
Fe.41 Unfortunately, in this particular case of the present Y-
TZP/Ni nanocomposite, it is impossible to estimate by how
much the theoretical Ms-value was reduced because of the
size effect, since we are dealing with a combination of two
phenomena.
30
20
10
0
Y-TZP/Ni nanocomposite
Y-TZP-0.3mol% NiO
solid solution
− 10
− 20
− 30
(6) Magnetic Coercivity and Size of Ni Precipitate
The magnetic coercivity of the Y-TZP/Ni nanocomposite deter-
mined from the magnetization curve (Fig. 4) was 0.81 kA/m,
which was apparently 10 times higher than the values for de-
magnetized bulk Ni, 0.08 kA/m.43 Indeed, the coercive force of
the ferromagnetic material increases with decrease in the particle
size and accompanies magnetic structure transition from multi-
domain to single-domain state to minimize the total magnetic
energy.44 The coercivity drastically reduces when the particle
size becomes smaller than critical single-domain size, and the
ferromagnetic material exhibits super-paramagnetic behavior as
already reported by numerous researchers.45–50
− 800 − 600 − 400 − 200
0
200
400
600
800
Magnetic field (kA/m)
Fig. 4. Magnetization curves of yttria-stabilized tetragonal zirconia (Y-
TZP)–0.3 mol% NiO solid solution and Y-TZP/Ni nanocomposite after
the internal reduction treatment measured by the super-conducting
quantum interference magnetometer at room temperature.
Hence, Gong et al.48 determined the size of a single domain in
Ni particles (32 nm) from the coercivity measurements and
compared the result with calculated size (42 nm) according
to the magnetic domain theory.45 Further, Estournes et al.49
claimed that Ni particles with a critical diameter of 20 nm be-
have super-paramagnetically, based on theory by Kneller and
Luborsky.50
The Ni particles investigated in the present study were as
small as 20 nm (see the discussion of TEM results), the value
that is lower than the single-domain size while higher than the
super-paramagnetic particle size. Thus, merely a part of Ni par-
ticles in the Y-TZP/Ni nanocomposite might account for super-
paramagnetic behavior, which indicates that magnetic coercivity
of the Y-TZP/Ni nanocomposite is lower than that reported
earlier for ceramic/Ni nanocomposites with larger Ni particle
size.4,12
TZP–NiO solid solution proves that metallic Ni is certainly in
the form of precipitations in Y-TZP/Ni nanocomposite after the
internal reduction. The magnetization curve obtained for the Y-
TZP/Ni nanocomposite was replotted using data for the mon-
olithic Y-TZP in such a way by subtracting the effect of dia-
magnetism of the NiO-doped Y-TZP matrix. Furthermore,
saturation magnetization (Ms) of the Y-TZP/Ni nanocompos-
ite was obtained by the Arrott plot.39
Assuming that all the added NiO was in the form of metallic
Ni particles after reduction, the obtained Ms-value for the Y-
TZP/0.1 vol% Ni nanocomposite equaled 21.7 emu/g (magnet-
ization of a unit mass of metallic Ni). This was considerably
lower than the one reported previously for pure bulk nickel
(55 emu/g),40 while the observed difference is because of
the small amount of metallic Ni particles precipitated in the
nanocomposite.
It should be noted again that fracture toughness of the Y-
TZP–0.3 mol% NiO solid solution and its reduced product (Y-
TZP/Ni nanocomposite) is practically of the same value, which
indicated that solid solute NiO was not completely transformed
into metallic Ni particles in the Y-TZP/Ni nanocomposite after
the internal reduction process. These results well agree with the
magnetic measurements, which proved the low value of Ms for
the Y-TZP/Ni nanocomposite.
In order to evaluate the Ni content in the reduced sample, the
Ms-values for pure bulk Ni were compared with the measured
ones. Hence, it was found that the amount of metallic Ni in the
Y-TZP/Ni nanocomposite was equal to 39.5 at.%, while using
Ms-values of 21.7 and 55 emu/g for the nanocomposite and
nickel, respectively. It indicates that the amount of Ni precipi-
tate in Y-TZP is approximately 0.04 vol%, and hence the nano-
composite prepared in this investigation may be described as
(Y,Ni)-TZP/0.04 vol% Ni. In materials of this type dealt with
ceramics with embedded nano-sized metallic particles, the size
effect could not be ignored. Indeed, the low angle neutron scat-
tering study of nanostructured iron by Wagner et al.41 already
revealed that the interface of ferromagnetic particles has sig-
nificantly lower average magnetic moment than the grain inte-
rior. Consequently, they claimed lower saturation magnetization
of nanostructured Fe than that of bulk material. Moreover, a
similar research for nickel by Host et al.42 agrees with Wagner
et al.’s41 results.
IV. Conclusions
The Y-TZP/Ni nanocomposite can be fabricated by the modi-
fied internal reduction method, which includes solid solution—
precipitation process. Fracture toughness of the nanocomposites
produced by the indicated process was remarkably improved.
The effect was caused by a small amount of NiO (B0.3 mol%)
that was solid solute into the Y-TZP, while the Y-TZP/Ni nano-
composite was still destabilized by the remaining nickel solution
after the reduction treatment. The presented results were con-
firmed by Raman spectroscopy as well as TEM observations. In
particular, high-magnification TEM revealed the characteristic
microstructure of the Y-TZP/Ni nanocomposite, in which
metallic Ni particles size embedded within the Y-TZP grains
equaled approximately 20 nm. Precipitated Ni was found to
accompany small pores and flat planes surrounding the
nanoparticles, and pores consisted of low-energy crystal planes
of zirconia.
The Y-TZP/Ni nanocomposite fabricated by the internal re-
duction method exhibits ferromagnetic behavior coexistence
with high fracture toughness. Using the measured data of sat-
uration magnetization, it was found that 39.5 at.% of doped
nickel was in situ transformed into nickel nanoparticles during
the reduction process. These results indicate that the nanocom-
posite material obtained by the present internal reduction is, to
be exact, NiO-doped Y-TZP dispersed with 0.04 vol% of Ni

June 2005
Mechanical and Magnetic Properties of Y-TZP/Ni Nanocomposite
1473
22H. Schmalzried, ‘‘Oxide Solid Solutions and Its Internal Reduction Reac-
tions,’’ Ber. Bunsenges. Phys. Chem., 88 [12] 1186–91 (1984).
23M. Backhaus-Ricoult and C. B. Carter, ‘‘Mechanism of the Internal Reduc-
tion of (Mg,Cu)O,’’ J. Am. Ceram. Soc., 70 [10] c291–4 (1987).
24M. Backhaus-Ricoult and S. Hagege, ‘‘Metal–Ceramic Interfaces in Internally
Reduced (Mg,Cu)O,’’ Acta Metall. Mater., 40, S267–74 (1992).
25M. Backhaus-Ricoult, S. Hagege, A. Peyrot, and P. Moreau, ‘‘Internal Re-
duction of Chromium-Doped a-Alumina,’’ J. Am. Ceram. Soc., 77 [2] 423–30
(1994).
(i.e., (Y,Ni)-TZP/0.04 vol% Ni nanocomposite). Nevertheless,
the fact obtained in the present investigation serves as a proof
that the method was entirely successful in the fabrication of ce-
ramic/metal nanocomposite with multifunctional properties.
Acknowledgment
26M. Backhaus-Ricoult and A. Peyrot-Chabrol, ‘‘Diffusion-Induced Grain-
Boundary Migration During Internal Reduction of Chromium-Doped Al2O3,’’
Philos. Mag. A, 73 [4] 973–98 (1996).
The authors acknowledge Prof. R. Nowak for his useful discussion and
comments.
27
.
E. Ustundag, R. Subramanian, R. Dieckmann, and S. L. Sass, ‘‘In Situ
¨
Formation of Metal–Ceramic Microstructures in the Ni–Al–O System by Partial
Reduction Reactions,’’ Acta Metall. Mater., 43, 383–9 (1995).
28L. Rebecca, A. Everman, and R. F. Cooper, ‘‘Internal Reduction of an Iron-
Doped Magnesium Aluminosilicate Melt,’’ J. Am. Ceram. Soc., 86 [3] 487–94
(2003).
References
1K. Niihara, ‘‘New Design Concept of Structural Ceramics–Ceramic Nano-
composites,’’ J. Ceram. Soc. Jpn., 99, 974–82 (1991).
2T. Ohji, Y.-K. Jeong, Y.-H. Choa, and K. Niihara, ‘‘Strengthening and Tough-
ening Mechanisms of Ceramic Nanocomposites,’’ J. Am. Ceram. Soc., 81, 1453–60
(1998).
29S. P. S. Badwal, ‘‘Effect of Dopant Concentration on the Grain Boundary
and Volume Resistivity of Yttria-Zirconia,’’ J. Mater. Sci. Lett., 6, 1419–21
(1987).
3F. Wakai, Y. Kodama, S. Sakaguchi, N. Murayama, K. Izaki, and K. Niihara,
‘‘A Superplastic Covalent Crystal Composite,’’ Nature, 344, 421–3 (1990).
4T. Sekino, T. Nakajima, S. Ueda, and K. Niihara, ‘‘Reduction and Sintering of
a Nickel-Dispersed-Alumina Composite and Its Properties,’’ J. Am. Ceram. Soc.,
80, 1139–48 (1997).
30H. Kondo, T. Sekino, T. Kusunose, T. Nakayama, Y. Yamamoto, M. Wada,
T. Adachi, and K. Niihara, ‘‘Solid-Solution Effects of a Small Amount of Nickel
Oxide Addition on Phase Stability and Mechanical Properties of Yttria-Stabilized
Tetragonal Zirconia Polycrystals,’’ J. Am. Ceram. Soc., 86, 523–5 (2003).
31M. Yoshimura, ‘‘Phase Stability of Zirconia,’’ Ceram. Bull., 67, 1950–5
(1988).
5T. Kusunose, T. Sekino, Y.-H. Choa, and K. Niihara, ‘‘Machinability of Sil-
icon Nitride/Boron Nitride Nanocomposites,’’ J. Am. Ceram. Soc., 85 [11] 2689–95
(2002).
32K. Niihara, R. Morena, and D. P. H. Hasselman, ‘‘Evaluation of KIc of Brittle
Solids by the Indentation Method with Low Crack-to-Indent Ratios,’’ J. Mater.
Sci. Lett., 1, 13–6 (1982).
6T. Nakayama, B. Skarman, L. R. Wallenberg, T. Sekino, Y.-H. Choa, T. A.
Yamamoto, and K. Niihara, ‘‘Characterization and Optical Properties of CeO2
Based Nanocluster Composites,’’ Scr. Mater., 44, 1929–32 (2001).
7M. Nawa, K. Yamazaki, T. Sekino, and K. Niihara, ‘‘Microstructure
33T. Nose and T. Fujii, ‘‘Evaluation of Fracture Toughness for Ceramic
Materials by a Single-Edge-Precracked-Beam Method,’’ J. Am. Ceram. Soc., 71,
328–33 (1988).
and Mechanical Behaviour of 3Y-TZP/Mo Nanocomposites Possessing
a
34D. R. Clarke and F. Adar, ‘‘Measurement of the Crystallographically Trans-
formed Zone Produced by Fracture in Ceramics Containing Tetragonal Zirconia,’’
J. Am. Ceram. Soc., 65, 284–8 (1982).
Novel Interpenetrated Intragranular Microstructure,’’ J. Mater. Sci., 31,
2849–58 (1996).
8T. Nakayama, Y.-H. Choa, T. Sekino, and K. Niihara, ‘‘Powder Preparation
and Microstructure for Nano-Sized Metallic Iron Dispersed MgO Based Nano-
composites with Ferromagnetic Response,’’ J. Ceram. Soc. Jpn., 108, 781–4 (2000).
9S.-T. Oh, M. Awano, M. Sando, and K. Niihara, ‘‘Fabrication of Nano-
Sized Ni–Co Alloy-Dispersed Alumina Nanocomposite,’’ J. Mater. Sci. Lett., 17,
1925–7 (1998).
35Y. Kubota, M. Ashizuka, and H. Hokazono, ‘‘Elastic Modulus and Fracture
Toughness of CeO2-Containing Tetragonal Zirconia Polycrystals,’’ J. Ceram. Soc.
Jpn., 102, 175–9 (1994).
36T. R. Lai, C. L. Hogg, and M. V. Swain, ‘‘Evaluation of Fracture Toughness
and R-Curve Behavior of Y-TZP Ceramics,’’ ISIJ Int., 29, 240–5 (1989).
37D.-Y. Kim, S. M. Wiederhorn, B. J. Hockey, C. A. Handwerker, and
J. E. Blendell, ‘‘Stability and Surface Energies of Wetted Grain Boundaries in
Aluminum Oxide,’’ J. Am. Ceram. Soc., 77 [2] 444–53 (1994).
38M. Kitayama, T. Narushima, W. Craig Carter, R. M. Cannon, and
A. M. Glaeser, ‘‘The Wulff Shape of Alumina: I, Modeling the Kinetics of Mor-
phological Evolution,’’ J. Am. Ceram. Soc., 83 [10] 2561–71 (2000).
39A. Arrott, ‘‘Criterion for Ferromagnetism from Observations of Magnetic
Isotherms,’’ Phys. Rev., 108, 1394–6 (1957).
10M. Awano, M. Sando, and K. Niihara, ‘‘Synthesis of Nanocomposite Ce-
ramics for Magnetic Remote Sensing and Actuating,’’ Key Eng. Mater., 161–163,
485–8 (1999).
11B.-S. Kim, J.-S. Lee, T. Sekino, Y.-H. Choa, and K. Niihara, ‘‘Hydrogen
Reduction Behavior of NiO Dispersoid During Processing of Al2O3/Ni Nano-
composites,’’ Scr. Mater., 44, 2121–5 (2001).
12H. Kondo, T. Sekino, Y.-H. Choa, T. Kusunose, T. Nakayama, M. Wada,
T. Adachi, and K. Niihara, ‘‘Mechanical and Magnetic Properties of Nickel Dis-
persed Tetragonal Zirconia Nanocomposites,’’ J. Nanosci. Nanotechnol., 2, 485–90
(2002).
40J. Crangle and G. M. Goodman, ‘‘The Magnetization of Pure Iron and Nick-
el;’’ pp. 477–91 in Proceedings of the Royal Society of London, Series A, Mathe-
matical and Physical Sciences, Vol. 321, Issue 1547, The Royal Society, London,
UK, 1971.
13T. Sekino and K. Niihara, ‘‘Microstructural Characteristics and Mechanical
Properties for Al2O3/Metal Nanocomposites,’’ Nanostruct. Mater., 6, 663–6
(1995).
41W. Wagner, A. Wiedenmann, W. Petry, A. Geibel, and H. Gleiter, ‘‘Magnetic
Microstructure of Nanostructured Fe, Studied by Small Angle Neutron Scatter-
ing,’’ J. Mater. Res., 6, 2305–11 (1991).
14R. A. Roy and R. Roy, ‘‘Dispersic Xerogels: I. Ceramic–Metal Composites,’’
Mater. Res. Bull., 19, 169–77 (1984).
15S.-T. Oh, M. Sando, T. Sekino, and K. Niihara, ‘‘Processing and Properties
of Copper Dispersed Alumina Matrix Nanocomposites,’’ Nanostruct. Mater., 10,
267–72 (1998).
42J. J. Host, J. A. Block, K. Parvin, V. P. Dravid, J. L. Alpers, T. Sezen, and
R. LaDuca, ‘‘Effect of Annealing on the Structure and Magnetic Properties
of Graphite Encapsulated Nickel and Cobalt Nanocrystals,’’ J. Appl. Phys., 83,
793–801 (1998).
16T. Sekino, J.-H. Yu, Y.-H. Choa, J.-S. Lee, and K. Niihara, ‘‘Reduction and
Sintering of Alumina/Tungsten Nanocomposites: Powder Processing, Reduction
Behavior and Microstructural Characterization,’’ J. Ceram. Soc. Jpn., 108, 541–7
(2000).
43G. W. C. Kaye and T. H. Laby, Tables of Physical and Chemical Constants,
16th edition; 175 pp, Longman, U.K., 1995.
44F. E. Luborsky, ‘‘Coercive Materials,’’ J. Appl. Phys., 32, 171S–83S (1961).
45C. Kittel, ‘‘Theory of the Structure of Ferromagnetic Domains in Films and
Small Particles,’’ Phys. Rev., 70, 965–71 (1946).
17R. Roy, ‘‘Ceramics by the Solution–Sol–Gel Route,’’ Science, 238, 1664–9
(1987).
18E. Breval, Z. Deng, S. Chiou, and C. G. Pantano, ‘‘Sol–Gel Prepared
Ni–Alumina Composite Materials. Part I. Microstructure and Mechanical Prop-
erties,’’ J. Mater. Sci., 27, 1464–8 (1992).
46W. F. Brown, ‘‘Relaxational Behavior of Fine Magnetic Particles,’’ J. Appl.
Phys., 30, 130S (1959).
47E. H. Frei, S. Shtrikman, and D. Treves, ‘‘Critical Size and Nucleation Field of
Ideal Ferromagnetic Particles,’’ Phys. Rev., 106, 446–55 (1957).
48W. Gong, H. Li, Z. Zhao, and J. Chen, ‘‘Ultrafine Particles of Fe, Co, and Ni
Ferromagnetic Metals,’’ J. Appl. Phys., 69, 5119–21 (1991).
49C. Estournes, T. Luts, J. Happich, T. Quaranta, P. Wissler, and J. L. Guille,
‘‘Nickel Nanoparticles in Silica Gel: Preparation and Magnetic Properties,’’
J. Magn. Magn. Mater., 173, 83–92 (1997).
19E. D. Rodeghiero, O. K. Tse, J. Chisaki, and E. P. Giannelis, ‘‘Synthesis and
Properties of Ni-a-Al2O3 Composites via Sol–Gel,’’ Mater. Sci. Eng. A, A195,
151–61 (1995).
20
.
R. Subramanian, E. Ustundag, S. L. Sass, and R. Dieckmann, ‘‘In-Situ For-
¨
mation of Metal–Ceramic Microstructures by Partial Reduction Reactions,’’ Solid
State Ionics, 75, 241–55 (1995).
21E. T. Trukdogman and J. V. Vinters, ‘‘Gaseous Reduction of Iron Oxides—
1,’’ Met. Trans., 2 [11] 3175–88 (1971).
50E. F. Kneller and F. E. Luborsky, ‘‘Particle Size Dependence of Coercivity and
Remanence of Single-Domain Particles,’’ J. Appl. Phys., 34, 656–8 (1963).
&