Professional Documents
Culture Documents
Abstract
The biocompatibility of zirconia–alumina (ZA) nano-composites in load-bearing applications such as dental/orthopedic implants
was significantly enhanced by the addition of bioactive HA. The ZA matrix was composed of nano-composite powder obtained
from the Pechini process and had higher flexural strength than conventionally mixed zirconia–alumina composite. Because the ZA
nano-composite powder effectively decreased the contact area between HA and zirconia for their reaction during the sintering
process, the HA-added ZA nano-composites contained biphasic calcium phosphates (BCP) of HA/TCP and had higher flexural
strength than conventionally mixed ZA–HA composite. From the in vitro test with osteoblastic cell-lines, the proliferation and the
differentiation (as expressed by the alkaline phosphatase activity) of the cellular response on the HA-added ZA nano-composites
gradually increased as the amount of HA added increased. From the mechanical and biological evaluations of the HA-added ZA
nano-composites, 30HA (30 vol% HA+70 vol% ZA) was found to be the optimal composition for load-bearing biological
applications.
r 2004 Elsevier Ltd. All rights reserved.
0142-9612/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2004.02.061
ARTICLE IN PRESS
510 Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517
and is classified as a bioinert ceramic [1]. Therefore, ZA formed in the gel as the polymer content increased, the
is considered to be one of the most suitable materials to zirconia gradually grew into nano-sized particles and
be used as a matrix for the HA-added strong composites. eventually became a nano-composite powder through
This study focused on improving the biocompatibility their merging with alumina.
of strong ZA nano-composites through the addition of
bioactive HA. The strength and biocompatibility of the 2.2. Preparation and characterization of HA-added
composites were measured as a function of the HA composites
content and correlated with the compositional and
structural variations. This study verifies the potential The starting powders were commercial HA (Alfa
of HA-added ZA nano-composites to be used in load- Aesar Co., Ward Hill, MA, USA) and the above-
bearing applications. mentioned ZA nano-composite powder. In addition,
commercially available ZrO2 (3Y-TZP, Tosoh Co.,
Tokyo, Japan) and Al2O3 (AKP50, Sumitomo Co.,
2. Materials and methods Osaka, Japan) were used for the purpose of comparison.
The ZA nano-composite powders (composition of
2.1. ZA nano-composite powder by Pechini process 80 wt% ZrO2–20 wt% Al2O3) were mixed with HA
(specific compositions of up to 40 vol%) by ball milling
The Pechini process is based on the formation of a for 24 h. After the mixed slurry was dried, crushed and
network which is composed of a polymeric precursor sieved, the resulting powders were hot-pressed in a
and a cationic network modifier [13]. In the Pechini graphite mold at 1400 C with an applied pressure of
process, the metallic ions are distributed homogeneously 30 MPa for 1 h. The phase of the hot-pressed bodies was
on an atomic scale throughout the polymeric network. analyzed with an X-ray diffractometer (M18XHF, MAC
Such a structure eliminates the need for long-range Science, Yokohama, Japan), and the density of the
diffusion during the subsequent formation of metal specimens was measured by the Archimedes method. The
oxides. Therefore, in conventional Pechini processes, the fractured and polished surfaces were observed with a
precursor forms a homogeneous single phase oxide of SEM (JSM-6330F, JEOL, Tokyo, Japan). The flexural
precise stoichiometry at a relatively low temperature. In strength was measured by means of the four-point flexure
this experiment, we used Zr and Al as the metallic ions test with a cross-head speed of 0.5 mm/min, and inner-
in the Pechini process to make a nano-composite and outer-spans of 20 and 40 mm, respectively (number
powder. Since zirconium and aluminum are unlikely to of samples=7) [8]. The tensile surface of the specimens
form a compound, the intimacy between the Zr and Al was polished using diamond slurries (1 mm finish).
ions captured in a polymeric network enables them to
form a nano-sized composite powder. 2.3. Biocompatibility
Metal chlorides (ZrCl2O 8H2O, AlCl3 6H2O,
YCl3 6H2O) were used as the cationic sources; citric The specimens used for the cell tests (proliferation and
acid monohydrate (C6H8O7 H2O, CAM) and ethylene differentiation) were prepared from the hot-pressed
glycol (C2H6O2, EG) were used as the polymeric matrix. specimens, so as to have a square shape with dimensions
A stoichiometric mixture of Zr and Y source solution of 10 mm 10 mm 1 mm. After polishing both sides
(3 mol% Y2O3 doping in ZrO2) and Al source solution with diamond slurries down to 1 mm, the specimens were
were used as the starting materials. The compositional washed and sterilized at 121 C for 20 min.
ratio of ZrO2 to Al2O3 was 4:1 by weight, because this
composition showed the highest flexural strength [12].
After mixing the metallic source solutions with CAM-
EG solution, the resultant mixture was heated to 130 C
to promote esterification between CAM and EG. Finally,
the gel was dried, crushed and calcined at 800 C. The
obtained powders were ball-milled in ethanol for 48 h
with ZrO2 balls to reduce the secondary particle size.
The total polymer content in the powder precursor
was as high as 90 wt%. When 90% of polymer was
added to the solution, the particles began to coalesce so
as to form a nano-composite powder composed of ZA.
All of the particles were composed of 100–200 nm (A) 50 nm (B) 50 nm
clusters, which contained zirconia nano-crystallites
whose size was about 10 nm, as shown in Fig. 1(A) Fig. 1. TEM images of ZrO2–Al2O3 nano-composite powder; (A)
and (B). Since a greater number of nucleation sites were bright field image and (B) dark field image.
ARTICLE IN PRESS
Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517 511
In order to investigate the proliferation behavior on the Fig. 2(A) shows the ZA nano-composite without HA, in
ZA–HA composites, MG63 cell lines were cultured in a which dark alumina grains were homogeneously dis-
dish containing Dulbecco’s modified Eagle’s medium persed in a bright zirconia matrix. The homogeneous
(DMEM, Invitrogen Co., Carlsbad, CA, USA) supple- microstructure and very small grain size (about 250 nm)
mented with 10% fetal bovine serum (FBS, Invitrogen are well illustrated in this micrograph. When 10 vol%
Co., Carlsbad, CA, USA). The fabricated specimens, HA was added, the microstructure was changed, as
including pure HA as a positive control, were placed into a shown in Fig. 2(B); the small dark grains of alumina
24-well plate, which was followed by plating 3 104 cells/ became coarser, and large dark grains (B1 mm) of HA
cm2 onto the discs. The cells were cultured for 7 days in a appeared in the microstructure. When 20 and 30 vol%
humidified incubator under an atmosphere of 5% CO2 at HA were added (Figs. 2(C) and (D)), the resulting
37 C. After detaching the cells from the specimens with a alumina grains had an acicular shape, while the HA
0.05% trypsin-EDTA solution and staining with 0.4% grains were plate-like in shape. The pores (about the
trypan blue, the living cells were counted using a same size as the zirconia grains) were thought to be
hemocytometer (Paul Marienfeld GmbH & Co., KG, formed during the polishing and thermal etching
Lauda-K.onigshofen, Germany). Each set of tests was processes.
performed in triplicate, and the data were normalized so Fig. 3 shows the densities of the ZA nano-composites
as to take the surface area into consideration. For the containing various amounts of HA. Compared to the
SEM observations, the 5 day-cultured cells (1 104 cells/ theoretical density, which was calculated by the rule of
cm2) were fixed with 2.5% glutaraldehyde, and then mixture, the density of the specimens with a low HA
dehydrated with graded ethanol (70%, 90%, and 100%). content almost matched with the theoretical value.
In order to measure the differentiation level expressed However, as the amount of HA was increased up to
by the alkaline phosphatase (ALP) activity of the HOS 40 vol%, the discrepancy between the measured and
cell-line on the specimens, the cells were plated at a theoretical densities increased.
density of 1 104 cells/cm2 in a 24-well plate, and then The composition of the specimens was analyzed
cultured for 10 days. At harvest, the culture media were by means of their XRD patterns. When no HA was
decanted, and the cell layers were washed once with added, only peaks corresponding to ZrO2 and Al2O3
HBSS, followed by trypsinization. After centrifugation were detected, as shown in Fig. 4(A). With the
at 1200 rpm for 7 min, the cell pellets were washed with addition of 20% and 30% HA (Figs. 4(B) and (C),
PBS and resuspended by shaking them in 100 ml of 0.1% respectively), as well as the peaks corresponding to HA,
Triton X-100. The pellets were disrupted by 5 cycles of peaks were also detected for TCP, which originated
sequential freezing/thawing. After centrifugation at from the reaction between HA and ZrO2. These XRD
13,000 rpm for 20 min at 4 C, the cell lysates were patterns illustrate that sintering the composite at 1400oC
quantified using the BioRad DC protein assay (Bio-Rad led to the appearance of BCP of HA/TCP in the
Laboratories Inc., Hercules, CA, USA) and assayed specimen.
colorimetrically for ALP activity using p-nitrophenyl The flexural strengths of the HA-added ZA nano-
phosphate as the substrate (Sigma, St. Louis. MO, composites are shown in Fig. 5. The strength of the pure
USA). Each reaction was initiated with p-nitrophenyl ZA nano-composite was 1460 MPa when it was hot-
phosphate and lasted for 10 min at 37 C, before being pressed at 1400 C. When 5 vol% of HA was added to
stopped by quenching on ice. The p-nitrophenol the ZA nano-composite powder and it was hot-pressed
produced was measured at 410 nm using a spectro- under the same conditions, the flexural strength
photometer (UV-1700, Shimadzu Co., Kyoto, Japan). decreased to 1270 MPa. As the HA content was
increased up to 40 vol%, the flexural strength decreased
2.4. Statistical analysis steadily (ANOVA, po0.01). The hot-pressed specimens
obtained from the conventionally mixed ZA powder
A one-way analysis of variance (ANOVA) was used showed a similar trend in terms of their strength (open
to compare the means of the different groups, and circle in Fig. 5) with respect to the addition of
statistical significance was accepted at the 0.05 con- HA, however, their strengths were B than those of
fidence level. B powders.
Fig. 6 shows the fracture surfaces of the specimens.
When the pure ZA nano-composite was hot-pressed
3. Results at 1400 C, the fracture proceeded mostly in an
intergranular pattern and the fracture surface was
3.1. Evaluation of HA-added ZA nanocomposites very rough, as shown in Fig. 6(A). When 20 vol%
of HA was added to ZA nano-composite (Fig. 6(B)),
Figs. 2(A)–(D) show the microstructures of the ZA the fracture also occurred in an intergranular
nano-composites containing different amounts of HA. mode. However, the fracture morphology was slightly
ARTICLE IN PRESS
512 Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517
(A) (B)
HA
(C) (D)
HA
HA
Fig. 2. SEM images of ZA (80 wt%ZrO2–20 wt% Al2O3) composites; (A) without HA, (B) containing 10 vol% HA, (C) containing 20 vol% HA, and
(D) containing 30 vol% HA. All specimens were hot-pressed at 1400 C for 1 h, followed by thermal etching at 1300 C. The bright and dark grains in
the image belong to zirconia and alumina, respectively.
5.6
Z
(A) Z
A A
Z A
5.2
Intensity [Arb. Unit]
Density [g/cm ]
3
4.4
4.0 (C)
0 20 40
25 30 35 40 45
HA content [vol%] 2θ [o]
Fig. 3. Density of ZA–HA composites hot-pressed at 1400 C as a Fig. 4. XRD data of the ZA composites; (A) without HA, (B)
function of the HA content. containing 10 vol% HA, and (C) containing 30 vol% HA. All
specimens were hot-pressed at 1400 C: () HA, (~) TCP, (A)
alumina, and (Z) zirconia.
different, due to the large size of the HA grains. 3.2. Biological evaluations
In the case of pure HA (hot-pressed at 1200 C),
the fracture morphology was much different from SEM images of the MG63 cells which were cultured
that of the other specimens, with the fractures for 5 days on various specimens are shown in Fig. 7; (A)
mostly occurring a transgranular pattern, as shown in pure ZA nano-composite, (B) 10 vol% HA+90 vol%
Fig. 6(C). ZA (10HA), (C) 30 vol% HA+70 vol% ZA (30HA),
ARTICLE IN PRESS
Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517 513
and (D) pure HA. The cells on all specimens evolved 10% HA was added, the cells proliferated more actively
stellately and, in each case, the cells came into contact on the composite as compared to that on the pure ZA
with each other through the elongation of the cytosols. (ANOVA, po0.05). As the HA content increased, the
There seemed to be no significant morphological proliferation rate increased steadily (ANOVA,
differences between the various ZA nano-composites po0.001). Interestingly, when 40% HA was added to
containing different amounts of HA. the ZA, the proliferation rate on the corresponding
The number of cells on each specimen after prolifera- specimen was comparable with that on the pure HA
tion for 7 days was plotted, and the resulting graph is specimen.
shown in Fig. 8. The cells proliferated actively on all of Figure 9 shows the differentiation behavior expressed
the specimens including on pure ZA. However, when in terms of the ALP activity of the HOS cell lines on the
specimens after 10 days culture. The ALP activity
exhibited a similar trend to that of the proliferation.
2000
(ANOVA, p < 0.01) However, in contrast to the proliferation behavior, the
specimen containing 10% HA had a similar expression
1600 level compared to the pure ZA (ANOVA, p=0.062).
Flexural Strength [MPa]
0 20 40 4. Discussion
HA content [vol%]
Fig. 5. Flexural strength of the composites containing various As a bioinert implant material, ZA nano-composite is
amounts of HA. The error bars represent 71 standard deviations. considered to have sufficient flexural strength, however
The mean values are significantly different (ANOVA, po0.01). its biocompatibility has to be improved by the addition
(A) (B)
(C)
Fig. 6. Fractured surface images of the specimens; (A) pure ZA, (B) ZA containing 20 vol% HA, and (C) pure HA.
ARTICLE IN PRESS
514 Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517
Fig. 7. Morphologies of proliferated MG63 cells on (A) pure ZA, (B) 10HA (10 vol% HA+90 vol% ZA) composite, (C) 30HA (30 vol%
HA+70 vol% ZA) composite, and (D) pure HA after 5 days culture.
10
ALP Activity [µmol p-nitrophenol/h/mg protein]
8 (p < 0.05)
5
0.8 (p = 0.062)
( p < 0.05)
6
0.6
4 0.4
2 0.2
0.0
0 ZA 10HA 20HA 30HA 40HA HA
ZA 10HA 20HA 30HA 40HA HA
Specimen
Specimen
Fig. 9. Differentiation level as expressed by the ALP activity on the
Fig. 8. Proliferation results on the specimens after 7 days culture. The specimens after 10 days culture. The ALP expression level of the ZA–
number of cells was significantly increased by the addition of 10% HA HA composite was significantly different from that of pure ZA when
to the ZA (ANOVA, po0.05). The number of cells increased steadily 20% HA was added (ANOVA, po0.05). The expression level
with further addition of HA (ANOVA, po0.001). When 40% HA was increased steadily with increasing HA content (ANOVA, po0.001).
added to the ZA, the proliferation rate on that specimen became The expression level of the composite became comparable with that of
comparable to that on the pure HA. pure HA when 40% HA was added. (ANOVA, p>0.05).
of calcium phosphates such as HA or TCP, in order for ceramics. The reaction products, such as tricalcium
satisfactory osseointegration and faster bone regenera- phosphate [TCP; Ca3(PO4)2], dissolve in the body fluid
tion to be achieved [14,15]. much faster than HA. As a bioresorbable material, TCP
When HA-based composites are made using other might be ideal for some biomedical applications [1].
ceramics as reinforcing agents, chemical reactions have However, as a constituent for the composite material for
been found to occur between the HA and the other load-bearing applications, the use of TCP is not
ARTICLE IN PRESS
Y.-M. Kong et al. / Biomaterials 26 (2005) 509–517 515
400 0.6
Strength
Acknowledgements
0.5
0
This work was supported by a grant from the Korea
0 20 40 60 80 100 Health 21 R&D Project, the Ministry of Health &
HA content [vol%] Welfare, Republic of Korea (02-PJ3-PG6-EV11-0002).
Fig. 12. Effect of HA addition on the flexural strength and bioactivity
(ALP activity) of the ZA composites. The error bars represent 71
standard deviations (for flexural strength) and 71 standard errors (for
ALP activity). References
[15] Monma H. Tricalcium phosphate ceramics complexed with [19] Huang J, Di Silvio L, Wang M, Tanner KE, Bonfield W. In vitro
hydroxyapatite. J Ceram Soc Jpn 1987;96:60–4. mechanical and biological assessment of hydroxyapatite rein-
[16] Alam MI, Ashhina I, Ohmamiuda K, Enomoto S. Comparative forced polyethylene composite. J Mater Sci Mater Med 1997;8:
study of biphasic calcium phosphate ceramics impregnated with 775–9.
rhBMP-2 as bone substitutes. J Biomed Mat Res 2001;54:129–38. [20] Dalby MH, Di Silvio L, Harper EJ, Bonfield W. In vitro
[17] Kurashina K, Kurita H, Wu Q, Ohtsuka A, Kobayashi H. evaluation of a new PMMA cement reinforced with hydroxyapa-
Ectopic osteogenesis with biphasic ceramics of hydroxyapatite tite. J Mater Sci Mater Med 1999;10:793–6.
and tricalcium phosphate in rabbits. Biomaterials 2002;23:407–12. [21] Li J, Hastings GW. Oxide bioceramics: inert ceramic materials in
[18] Moursi AM, Winnard AV, Winnard PL, Lannutti JJ, Seghi RR. medicine and dentistry. In: Black J, Hastings GW, editors.
Enhanced osteoblast response to a PMMA-HA composite. Handbook of biomaterials properties. London: Chapman & Hall;
Biomaterials 2002;23:133–44. 1998. p. 340–54.