Acta Biomaterialia 3 (2007) 896–904

www.elsevier.com/locate/actabiomat

Bioinspired structure of bioceramics for bone regeneration

in load-bearing sites
Faming Zhang a, Jiang Chang a,*
, Jianxi Lu a,b
, Kaili Lin a, Congqin Ning a

a
Biomaterials and Tissue Engineering Research Center, Shanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 200050, China
b
Shanghai Bio-lu Biomaterials Co. Ltd., Shanghai 200335, China

Received 18 January 2007; received in revised form 11 May 2007; accepted 15 May 2007
Available online 26 June 2007

Abstract

The major problem with the use of porous bioceramics as bone regeneration grafts is their weak mechanical strength, which has not
been overcome to date. Here we described a novel way to solve this problem. Beta-tricalcium phosphate (b-TCP) bioceramics with a
bioinspired structure were designed and prepared with a porous cancellous core (porosity: 70–90%) inside and a dense compact shell
(porosity: 5–10%) outside that mimics the characteristics of natural bone. They showed excellent mechanical properties, with a compres-
sive strength of 10–80 MPa and an elastic modulus of 180 MPa–1.0 GPa, which could be tailored by the dense/porous cross-sectional
area ratio obeying the rule of exponential growth. The in vitro degradation of the bioinspired bioceramics was faster than that of dense
bioceramics but slower than that of porous counterparts. The changes in mechanical properties of the bioinspired ceramics during
in vitro degradation were also investigated. A concept of the bioinspired macrostructure design of natural bone was proposed which
provided a simple but effective way to increase the mechanical properties of porous bioceramics for load-bearing bone regeneration appli-
cations. It should be readily applicable to other porous materials.
Ó 2007 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Bioinspired structure; Bioceramics; Calcium phosphate; Load-bearing sites; Bone regeneration

1. Introduction with a porous form have very low mechanical strength,

With the aging of populations and prolonged life expec-
tancy, there is an increasing demand for bone grafts or syn-
thetic materials that can potentially replace, repair or
regenerate lost, injured or diseased bone [1]. Beta-trical-
cium phosphate (b-TCP) bioceramics have been extensively
explored as grafts for bone regeneration due to their good
biocompatibility and biodegradability. Bioceramics with a
porous structure that mimics the morphology of cancellous
bone are required for better osteoconductive capacity, and
this porous network could promote cell attachment, prolif-
eration and differentiation, provide pathways for biofluids
and allow tissue ingrowth [2]. However, b-TCP bioceramics
*
Corresponding author. Tel.: +86 21 5241 2804; fax: +86 21 5241 3903.
E-mail address: jchang@mail.sic.ac.cn (J. Chang).
which has restricted their use to low- or non-load-bearing
applications. Even though many attempts have been used
to increase their mechanical strength for load-bearing
applications, including introducing a secondary phase
[3–5], refining the grain size [6,7] and optimizing sintering
techniques [8], the weak mechanical properties of porous
b-TCP bioceramics has not yet been overcome.
Natural bone has long been the object of biomimetic
study because of its characteristic structure and excellent
biomechanical properties. Bone is a non-stoichiometric
carbonated apatite/collagen with a nanocomposite compo-
sition. It is a complex biomineralized system with an intri-
cate hierarchical microstructure assembled through the
deposition of apatite minerals within a collagenous matrix
[9]. The macrostructure feature of bone is porous cancel-
lous bone inside with dense compact (or cortical) bone shell
outside. This feature provides excellent biomechanical

1742-7061/$ - see front matter Ó 2007 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.actbio.2007.05.008
 F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904
properties that are modulated by the compact/cancellous
bone ratio. In the past several years, biomimetic studies
of natural bone have focused mainly on composition
mimesis, such as HA/collagen composites [10–12], micro-
structure mimesis, such as artificial bone materials with a
bone-like microstructure [9,13,14], or process mimesis
using self-assembling synthesize bone scaffolds [15,16].
However, the macrostructural mimesis of natural bone
has received less attention.
Here the structure of natural bone has inspired us to
solve problems related to the application of porous
b-TCP bioceramics in load-bearing sites. To increase the
mechanical properties of porous bioceramics, macrostruc-
ture bioinspired b-TCP bioceramics that mimic the struc-
ture of natural bone were designed and fabricated. The
mechanical properties and in vitro degradation of the fab-
ricated bioceramics were evaluated. Additionally, a con-
cept of bioinspired macrostructure design of natural bone
is proposed and discussed.
2. Materials and methods
2.1. Materials
b-TCP powders were synthesized by chemical precipita-
tion reaction by dropping a Ca(NO3)2 solution into a stir-
red (NH4)2HPO4 solution to produce calcium phosphate
precipitates. The pH of the solution was adjusted using
NH3 Æ H2O. After aging and washing, the precipitates were
dried at 80 °C for 24 h. Finally, the synthesized powders
were calcined at 800 °C for 2 h. The mean particle size of
calcined b-TCP powders was about 1 lm and the composi-
tion was pure b-TCP phase characterized by scanning elec-
tron microscopy (SEM) and X-ray diffraction (XRD).
2.2. Preparation procedure
The bioinspired structural b-TCP bioceramics were
prepared by injected molding. Polystyrene microballs with
diameters of 400–600 lm were packed in a die and then
heat-treated at 80 °C for 15 min to form the template.
The b-TCP powders along with dispersants and surfac-
tants were dispersed in distilled water to make ceramic
slurries with about 60% solid loading. The b-TCP slurries
were then injected into the pre-prepared template. The
green samples were heat-treated at 200 °C for 10 h to
drive off any organic substances, and then sintered at
800–900 °C for 2 h at a heating rate of 1 °C min1. For
a comparison, porous structure was also fabricated by
the polymer sponge method. Polyurethane foam cut into
desired shapes and sizes was immersed into the b-TCP
slurries until the internal pores were filled with ceramic
slurry. The impregnated sponge was passed through roll-
ers to remove the excess suspension. These samples were
dried in air for 24 h, sintered at 600 °C for 2 h to burn
out the polyurethane foams and then sintered at 800–
900 °C for 2 h at a heating rate of 1 °C min1. Next, the
897
two types of porous samples were placed in the center
of the die. Then the b-TCP ceramic slurries were injected
around the porous structure to form a dense shell. After
drying in air for 24 h, the bioinspired samples were
sintered at 1050–1150 °C for different time at a heating
rate of 1 °C min1, and then furnace cooled to room
temperature.
2.3. Microstructure characterization
X-ray diffraction (Rigaku D/max 2550V, Japan) with
Cu Ka radiation was used to characterize the phase com-
position of the bioceramics. The microstructure analysis
of the prepared bioceramics samples was performed by
SEM (JSM-6700F, Japan). The porosity of the dense, por-
ous and bioinspired bioceramics were measured by Archi-
medes’ method. A porcine femur was calcined at 900 °C
for 2 h, then subjected to an electron microprobe analysis
(8705QH2, Japan).
2.4. Measurement of mechanical properties
The compressive strength was conducted on an Instron
(5566, USA) testing machine with a cross-head speed of
0.5 mm min1. The elastic modulus was calculated from
the slope of the initial linear portion of the stress–strain
curve. For mechanical strength, three samples of each
group were used for statistical analysis.
2.5. Degradation in vitro
The in vitro degradation of the porous, bioinspired and
dense bioceramics was evaluated in a Tris–HCl buffer solu-
tion. The Tris–HCl solution was prepared by dissolving
tris–hydroxymethyl aminomethane (analytical reagent
grade) in distilled water with buffering at pH 7.4 ± 0.1 by
1.0 mol l1 of hydrochloric acid at 37 °C. The ratio of the
bioceramic weight (g) to solution volume (ml) was 5/100.
The bioceramics were suspended in polystyrene bottles
containing Tris–HCl solution at 37.0 °C in a 2 Hz fre-
quency-shaking bath. Three samples of each group were
collected after 120 h of immersion. Subsequent to immer-
sion, the solutions were vacuum filtered. The Ca and P con-
centrations in the filtered solutions were measured by
inductively coupled plasma-atomic emission spectrometer
(Vista AX ICP-AES). The samples and filtrated particu-
lates were rinsed with deionized water, and then dried in
an oven at 150 °C for 48 h. Finally, the weight of the dried
sample and particulates was measured in a precision elec-
tronic balance, and the weight loss percentage of the bioce-
ramics calculated. Three bioinspired samples of each group
were immersed in Tris–HCl solution under the above-men-
tioned conditions for 1, 3, 5, 7, 14, 21 and 28 days without
refreshing the solution. At each time point, the samples
were removed and rinsed with deionized water. The
mechanical properties of the bioinspired bioceramics after
898 F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904

immersion were measured in the wet state at room
temperature.
3. Results and discussion
3.1. Modeling
Bone is composed of compact (cortical) and cancellous
(trabecular) materials. Compact bone is the dense outer
shell of bone, and has a low surface area, which makes
up a large portion of the skeletal mass. Cancellous bone
is an interior meshwork of trabecula, and has a relatively
high surface area but forms a smaller portion of the skele-
ton. Compact bone is stiffer than cancellous bone. A sim-
plified model of the natural bone macrostructure is
illustrated in Fig. 1.
A number of interrelated variables determine the bone
strength, including bone mineral density, bone quality
and bone geometry. The latter consists of cross-sectional
areas, etc. A physical assumption proposed here is that
the proportion of compact bone determines the mechanical
properties of natural bone, ignoring every other factor.
Therefore, the biomechanical strength of bone could be tai-
Compact
bone
o 850 °C, as determined from Fig. 2, and then the dense
Cancellous
bone
Fig. 1. The simplified model of bone macrostructure.
lored by changing the proportion of compact bone. An
equation from the model in Fig. 1 and the physical assump-
tion could be formulated as
 
S dense
r¼f ð1Þ
S porous
   2 
S dense R  r2
r¼f ¼f ð2Þ
S porous r2
where Sdense is the cross-sectional area of the compact
bone, Sporous is the cross-sectional area of the cancellous
bone, r is the biomechanical strength of bone, R is the ra-
dius of the entire bone and r is the radius of the cancellous
bone. As shown in Eq. (1), the mechanical strength of bone
with the value of Sdense/Sporous follows a kind of functional
relationship. For round cross-sectional bone grafts, the
mechanical strength follows a kind of functional relation
with R2/r2 value (Eq. (2)).
3.2. Preparation and microstructures
In the preparation of bioinspired b-TCP bioceramics,
the major problem is the interface adhesion between the
dense and porous parts due to their different shrinkage
rates during the sintering process. The porous b-TCP
bioceramics exhibited a much higher shrinkage rate than
their dense counterparts in the radial and axial directions
(Fig. 2). The porous bioceramics showed about 23% and
14% shrinkage in the radial and axial directions, respec-
r tively, whereas the dense bioceramics showed about 17%
and 8% shrinkage. To avoiding the shrinkage differences,
R the porous b-TCP bioceramics were pre-sintered at
bioceramics were injected surrounding the porous ceram-
ics. Finally, the dense/porous composites were pressureless
sintered at 1050, 1100 and 1150 °C for 2 h.
The phase compositions of the bioinspired bioceramics
sintered at different temperatures were analyzed by XRD
(Fig. 3). TCP has three polymorphs: b-TCP is stable below
1180 °C, a-TCP between 1180 and 1400 °C, and a 0 -TCP
above 1470 °C [17]. It can be seen from Fig. 3 that the

a 24
b 18
22
Porous β -TCP 16 Porous β -TCP
20
Dense β -TCP
18 Dense β -TCP 14
Shinkage (%)

Shinkage (%)

16 12
14
10
12
8
10
8 6
6 4
4
2
2
0 0

-2 -2
600 700 800 900 1000 1100 600 700 800 900 1000 1100

Temperature (°C) Temperature (°C)

Fig. 2. The radial (a) and axial (b) shrinkage rates of the porous and dense b-TCP bioceramics.

β -TCP
Intensity
10 20 30 40 50
2theta (deg.)
Fig. 3. The XRD pattern of the prepared bioinspired b-TCP bioceramics.
b-TCP bioceramics were stable from 1050 to 1150 °C, and
the original b phase was preserved without transformation
into a-TCP phase. The a-TCP bioceramics, though bioac-
tive, have proven to be less useful for bone regeneration
due to their excessively higher resorption rate than that
of the b-TCP. Moreover, the volume change associated
with the b-TCP to a-TCP transformation often results in
the formation of microcracks that weaken the material.
Neither a-TCP phase nor impurities were detected in the
XRD pattern of the sintered bioceramics. However, the
b-TCP ceramics were not densified at 1050 °C, with many
micropores being left. The higher temperature (1150 °C)
sintering resulted in a coarser grain size. The optimal sin-
tering temperature of the bioinspired b-TCP bioceramics
was therefore determined in this study to be 1100 °C.
Fig. 4 shows a prepared bioinspired b-TCP sample that
was sintered at 1100 °C. The inner porous structure mimics
cancellous bone and the outer-shelled dense structure mim-
ics compact bone.
Compact bone
Cancellous bone
Fig. 4. The prepared bioinspired b-TCP sample.
F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904 899
Fig. 5 shows SEM micrographs of bioinspired b-TCP
bioceramics prepared at 1100 °C. The interface between
the dense and porous parts of the bioinspired b-TCP bioce-
ramics was compact and tight (Fig. 5a), which is very
important in determining the properties of the dense/
porous bioinspired compounds. In the porous core, the
o
macropore size was about 500 lm in diameter (Fig. 5b).
(c) 1150 C The macropores were highly interconnective, and the diam-
eter of the interconnected pores was about 100 lm
o
(b) 1100 C (Fig. 5b). The porosity of the porous core and dense shell
was about 72% and 8%, respectively. The average porosity
o
(a) 1050 C of the whole bioinspired bioceramics was about 56%. The
interconnectivity was more than 95%. The microstructure
of the macroporous wall is shown in Fig. 5c. There were
60 70 80
some micropores with a diameter of 1 lm distributed uni-
formly in the pore walls. As the results show, the designed
bioceramics has both macroporous and microporous
pores. In contrast, the microstructure of the dense shell
of the bioinspired bioceramics shows fewer and smaller
micropores (Fig. 5d).
For a comparison, porcine femur was calcined and sub-
jected to SEM observation, as presented in Fig. 6. The
interface of compact/cancellous bone was tight and natural
(Fig. 6a). The porous cancellous bone had a spumous
structure, with a diameter of 300–600 lm (Fig. 6b). The
porous wall of the cancellous bone shows a high degree
of orientation, with micropore size of about 1 lm
(Fig. 6c). The microstructure of the compact bone shows
that the grain size was refined, and the micropores were
fewer and smaller than in the cancellous bone (Fig. 6d).
There are many similarities in the macrostructure com-
pared with our prepared bioinspired bioceramics in Fig. 5.
To further mimic the cancellous structures, the porous
parts of the bioinspired bioceramics were fabricated by
the polymer sponge method. However, the sponge scaffolds
showed very low compressive strength (400–600 kPa). The
porous bioceramic scaffolds prepared by packing micro-
balls as a replica template showed a compressive strength
of about 2 MPa. The materials prepared by the infiltration
of polymer sponges are weaker because the struts of the
reticulated structure are often cracked during pyrolysis of
the polymeric template, the macroporous walls are thin
and the macropores have irregular shapes. These problems
are avoided in the porous scaffolds prepared by packing
microballs as a replica template. The porous b-TCP scaf-
folds prepared by this latter method have proved effica-
cious as bone regeneration and tissue engineering
scaffolds [18]. Although the porous structure of the bioin-
spired bioceramics in Fig. 5 was different from that of the
natural cancellous bone, the pores size, interconnectivity
and interconnective diameter of the bioinspired bioceram-
ics could be designed, modulated and fabricated.
3.3. Mechanical properties
Fig. 7 shows a typical stress–strain curve of bioinspired
b-TCP bioceramics during the compressive strength test.
900 F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904

Fig. 5. The microstructures of the bioinspired b-TCP bioceramics showing the dense/porous interface (a), macroporous structure (b), macropore wall (c)
and dense compact bone (d).

Fig. 6. The microstructures of natural bone showing the compact/cancellous interface (a), macroporous structure (b), macropore wall (c) and compact
bone (d).

The stress increased sharply as a function of the strain the compressive extension of 1.33 mm, the bioinspired
increment at the early stage and decreased rapidly after bioceramic fractured. It showed a long ‘‘plateau’’ with a
the maximum compressive strength point of 63 MPa. At continuous compressive load after the sample reached its

70
60
Compressive stress (MPa)
50
40
30
20
10
0 ceramics, r0 is the strength of porous ceramics, Sdense/Sporous
-10
-0.5
Fig. 7. A typical stress–strain curve of the bioinspired b-TCP bioceramics.
peak strength. At the extension of 3.5 mm, the load stress
was increased but did not return to zero because the frac-
tured fragments were compressed once again. This is a typ-
ical characteristic of the brittle fracture of bioceramics. The
variation of the compressive strength and elastic modulus
of the bioceramics with different dense/porous cross-
sectional area ratio (Sdense/Sporous) is illustrated in Fig 8.
The compressive strength increased from 10 to 80 MPa
with an Sdense/Sporous ratio from 0.1 to 4.8, i.e., the dense
shell ranging from 1 to 6.5 mm with a settled porous core,
obeying the rule of exponential growth. In addition, the
elastic modulus increased from 180 MPa to 1 GPa with
the Sdense/Sporous ratio increment, again following exponen-
tial growth. However, the compressive strength of pure
porous bioceramics was only about 2.0 MPa and the elastic
modulus was about 20 MPa with a porosity of 72%, as
indicated in Fig. 8. The mechanical properties have been
improved by about 5- to 40-fold by the bioinspired struc-
ture design. As expected, the mechanical properties of the
bioinspired bioceramics were tailored by the different
dense/porous cross-sectional area ratios, which has verified
90
80
Compressive Strengh (MPa)
70
60
50
40
30
20
10
Porous
0 which are at least in the range of cancellous bone [21].
0 1
Fig. 8. The variation of the compressive strength and elastic modulus of
the bioinspired b-TCP bioceramics with dense/porous cross-sectional area
ratio.
F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904 901
the proposed physical assumption that the proportion of
compact bone determines the mechanical properties of
bone, and is consistent with the above-proposed equation
(1). The fitting of the mechanical values in Fig. 8 obey
the mathematical model deduced from the curve fitting
the mechanical results:
 
S dense =S porous
r¼ r0 exp ð3Þ
t
where r is the strength of the bioinspired dense/porous
is the dense/porous cross-sectional area ratio (0 < Sdense/
-1.0 -1.5 -2.0 -2.5 -3.0 -3.5 Sporous < 10) and t is related to the mechanical inherence
Extension (mm) and porosity of the whole ceramics. The functional relation-
ship in the proposed equations (1) and (2) is explained in
Eq. (3). The dependence of the compressive strength and
elastic modulus of the bioinspired dense/porous b-TCP
bioceramics follows the exponential law of dense/porous
mixtures, as described by Eq. (3), i.e., the function in Eqs.
(1) and (2) should be addressed as an exponential function.
The b-TCP bioceramics are intrinsically brittle and have
low mechanical strength, especially in the porous form.
The mechanical properties of porous b-TCP are related to
the porosity, pore size and shapes. A lower porosity and
smaller pore size could result in an increase in the mechanical
properties, but they are not of benefit for bone tissue in-
growths. Porous structures with regular pore shapes, e.g.,
the round pores in Fig. 5, showed much higher mechanical
strength than structures with irregular pores made from
the sponge method. The strength values obtained in this
study are through microstructure manipulation that is bioin-
spired structure design. The mechanism of the mechanical
property enhancement is due to the incorporation of a dense
shell outside of a porous core.
Porous materials always have poor mechanical proper-
ties. Applications of porous calcium phosphates in the
body have been limited by their low strength. Numerous
techniques [3–8] have been investigated to retain their use-
ful bioactive properties whilst providing more suitable
1400
mechanical properties for particular applications. How-
ever, these techniques are limited for the porous calcium
1200 phosphate used in load-bearing sites. In this study, excel-
Compressive strength
Elastic Modulus (MPa)
lent mechanical properties of the porous b-TCP bioceram-
Elastic modulus 1000
ics have been achieved by bioinspired structure design.
800 The average compressive strength of human cancellous
bone is 2–12 MPa [19]. Moreover, the compressive
600
strength of the human vertebral bone (load-bearing site)
400
ranges from 24 to 43 MPa, and femoral cancellous bone
(load-bearing site) is in the range of 48–80 MPa [20]. For
200 load-bearing bone defect applications, synthetic bone graft
materials should have suitable mechanical properties,
0
2 3 4 5
Consequently, a mechanical strength of bioinspired
Sdense/Sporous
b-TCP bioceramics in the range of 10–80 MPa is suitable
for load-bearing applications. In particular, their elastic
modulus in the range of 150–1000 MPa is much lower
than that of the compact bone (7–30 GPa) and compara-
902
ble to that of cancellous bone (50–500 MPa), so that stress
shielding of bone can be avoided for the lower elastic
modulus.
3.4. Degradation in vitro
The effect of the bioinspired structure on degradability
was studied in vitro by immersion in Tris–HCl buffer solu-
tion. As shown in Table 1, the porous, bioinspired and
dense b-TCP bioceramics were all degradable, showing
weight loss and Ca, P ions release after 120 h of Tris–HCl
immersion. The bioinspired b-TCP bioceramics showed
much higher weight loss and Ca, P ions release than the
dense bioceramics, but a little lower weight loss than the
porous bioceramics. The b-TCP bioceramics are a typical
degradable material whose degradability is regulated by
surface area, microporosity, crystallinity, grain size, etc.
The above phenomenon is mainly attributed to their differ-
ences in surface area. The slower degradation rate of the
dense shell in the bioinspired b-TCP bioceramics could bear
the load for a longer time. The porous core with a high
interconnective structure degraded faster, and could serve
as a scaffold for bone tissue ingrowth. The host tissues could
replace the porous b-TCP after 1–2 years. Due to the incor-
poration of the dense shell, it is predicted that natural host
tissues may be able to replace the bioinspired dense/porous
b-TCP bioceramics after more than 2 years implantation;
this needs to be proved by further in vivo investigations.
Table 1
Degradation in vitro of the porous, bioinspired and dense b-TCP
bioceramics
b-TCP Weight loss
(100%)
Porous 0.65 ± 0.03
Bioinspired 0.51 ± 0.03
Dense 0.16 ± 0.02
50
Compressive strength (MPa)
45
40
35
30
25
20
15
10
0 7
Immersion time (days)
Fig. 9. Wet-state mechanical properties of the bioinspired b-TCP bioce-
ramics in Tris–HCl buffer solution.
F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904
The mechanical properties of the bioinspired bioceram-
ics after immersion in Tris–HCl buffer solution were also
measured in the wet state. Fig. 9 shows the effect of the deg-
radation on the mechanical properties of the bioinspired
b-TCP bioceramics. The starting compressive strength of
the bioinspired samples with a shell thickness of 4 mm that
had been pre-wetted in deionized water for 60 min was
44.0 MPa in the wet state. In the first week, the compres-
sive strength of the bioinspired bioceramics dropped by
about 9.3%, from 44.0 to 39.9 MPa. The compressive
strength dropped by about 17% to 36.5 MPa and by
36.8% to 27.8 MPa after 2 and 3 weeks of immersion,
respectively. The strength of the bioceramics was reduced
by about 47.2% to 23.2 MPa after immersion for 4 weeks.
There is an obvious decrease in the wet-state mechanical
properties of the bioinspired b-TCP bioceramics during
the degradation process in the biomimetic solution. The
degradation mechanism of the Ca–P bioceramics with
good crystallization is mainly driven by dissolution, and
this dissolution occurs easily at the boundary of micro-
pores [22]. Therefore, the mechanical strength loss may
be due to the dissolution of some b-TCP grains occurring
at the boundary of the micropores, with the result that lar-
ger and more defects were generated in the bioceramics.
Understanding the mechanical properties of the bioin-
spired b-TCP bioceramics during the degradation process
is helpful for the effective prediction of the degradation
process. It will enable the design of better implants for par-
ticular clinical applications. The wet-state compressive
strength of the bioinspired b-TCP bioceramics after 4
weeks’ degradation is comparable to the low-end value of
vertebral bone (24–43 MPa [20]) and is still more than 11
times higher than the initial mechanical value of the porous
b-TCP bioceramics in the dry state (2 MPa). This indicates
Ca ion release P ion release
that the bioinspired dense/porous structure is effective in
(mg l1) (mg l)1
increasing the mechanical properties of porous bioceram-
29.9 ± 71.5 5.71 ± 1.5
ics. Shi et al. [23] reported that coating HA thin films
25.8 ± 31.2 4.64 ± 1.2
18.6 ± 61.2 1.30 ± 1.2 on reticulated alumina achieved compressive strengths of
7–10.35 MPa. The HA-coated porous alumina is bioactive
on the surface but not degradable, so can be used as a bone
substitute. Herein, the bioinspired structural b-TCP bioce-
ramics are degradable, and exhibited high compressive
strength before and after the degradation test. Therefore,
these bioceramics may be used for bone regeneration in
some load-bearing sites, e.g., human cervical vertebrae,
lumbar vertebrae and even femur.
Although we chose biodegradable ceramics b-TCP for
a bioinspired study in this study, it is believed that this
technique might be readily applicable to other porous
materials, including porous bioinert ceramics, bioactive
ceramics, bioglass, glass ceramics, metals, polymers and
their composites. Recently, Sylvain et al. [24] prepared by
freeze-casting HA scaffolds with a dense cellular–lamellar
14 21 28
structure in the longitudinal direction which exhibited com-
pressive strength of up to 65 MPa with 56% porosity; how-
ever, the processing technique is complicated and restricted
to freeze-casting. Kaito et al. [25] reported dual HA com-
 F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904
posites with porous and solid parts for use as bone substi-
tute materials, where the dense HA has 0% porosity and
the porous part has 75% porosity. The porosity of compact
bone ranges from 5% to 30%, while cancellous bone poros-
ity ranges from 30% to 90% [26]. Bone porosity is not fixed
and can change in response to altered loading, disease and
aging [26]. In this study, the biodegradable b-TCP bioce-
ramics with a bioinspired dense/porous structure mimic
the characteristics of natural bone for use as bone regener-
ation materials, where the dense b-TCP shell has 5–10%
porosity for considerable degradability and the porous core
has 70–90% porosity. The shell and core are in the range of
natural bone porosity. It is the bioinspired structure design
that enables these figures to be achieved. Our material pro-
cessing technique is also different from that of Kaito et al.
Herein, a concept of bioinspired macrostructure design of
natural bone is proposed. On early stage, dense forms of
biomaterials have been used to repair bone defects. For
ingrowth of bone tissue, porous biomaterials were explored
for clinical applications. From dense to porous form of
biomaterials was inspired from the cancellous bone struc-
ture. To increase mechanical properties of the porous bio-
materials, dense/porous biomaterials were explored here.
We predicted that the development tendency of bone repair
materials is from porous to dense/porous form, which
inspired from the macrostructure of bone. In this study,
the bioinspired bioceramics comprise a dense shell outside
and a porous core inside. Conversely, bioceramics can also
be prepared with a porous shell outside and a dense core
inside. The designs are all inspired by the macrostructural
features of natural bone. This bioinspired macrostructure
design of natural bone has provided a simple but effective
way to increase the mechanical properties of porous mate-
rials, and has resolved the problem of the weak strength of
porous materials in nature.
4. Conclusions
b-TCP bioceramics with a bioinspired structure were
designed and prepared that exhibit a similar macrostruc-
ture to natural bone, and the interface between the dense
and porous bioceramics is compact and tight. The bioin-
spired bioceramics revealed excellent mechanical proper-
ties, with compressive strength from 10 to 80 MPa and
elastic modulus from 180 to 1000 MPa, both of which are
5–40 times higher than that of the pure porous b-TCP
bioceramics. The degradability in vitro shows that the bio-
inspired b-TCP exhibit a higher degradation rate than the
dense b-TCP but a lower rate than the porous counterpart.
The changes in mechanical properties of the bioinspired
bioceramics during in vitro degradation were investigated
and showed 47.2% strength loss after 4 weeks of immer-
sion. A concept of bioinspired macrostructure design of
natural bone was proposed which provides a simple but
effective way to improve the mechanical properties of por-
ous materials.
903
Acknowledgements
The authors are grateful to Shanghai Postdoctoral Sci-
entific Key Program (No. 06R214201), China Postdoc-
toral Science Foundation (No. 20060390648), Science
and Technology Program of Shanghai Municipality (No.
05DZ05034) and State 973 Program of China (No.
2005CB522704) for the financial support. We also thank
Shanghai Bio-lu Biomaterials Co. Ltd. for the support
in the preparation of bioceramics samples.
References
[1] Tampieri A, Celotti G, Landi E. From biomimetic apatites to
biologically inspired composites. Anal Bioanal Chem 2005;381:568–76.
[2] Freyman TM, Yannas IV, Gibson LJ. Cellular materials as porous
scaffolds for tissue engineering. Progr Mater Sci 2001;46:273–82.
[3] Ramay HR, Zhang M. Biphasic calcium phosphate nanocomposite
porous scaffolds for load-bearing bone tissue engineering. Biomate-
rials 2004;25:5171–80.
[4] Zhang Y, Zhang M. Three-dimensional macroporous calcium phos-
phate bioceramics with nested chitosan sponges for load-bearing bone
implants. J Biomed Mater Res 2002;61(1):1–8.
[5] Samar JK, Himesh AB, Abhijeet D. MgO–Na2O–P2O5 based
sintering additives for tricalcium phosphate bioceramics. J Am
Ceram Soc 2006;89(3):875–81.
[6] Bow JS, Liou SC, Chen SY. Structural characterization of room-
temperature synthesized nano-sized b-tricalcium phosphate. Bioma-
terials 2004;25(16):3155–61.
[7] Prashant NK, Charles S, Lee DH, Dana O, Daiwon C. Nanostruc-
tured calcium phosphates for biomedical applications: novel synthesis
and characterization. Acta Biomater 2005;1:65–83.
[8] Wang XL, Fan HS, Xiao YM, Zhang XD. Fabrication and
characterization of porous hydroxyapatite/b-tricalcium phosphate
ceramics by microwave sintering. Mater Lett 2006;60(4):455–8.
[9] Liao SS, Cui FZ, Zhang W, Feng QL. Hierarchically biomimetic
bone scaffold materials: nano-HA/collagen/PLA composite. J Bio-
med Mater Res B 2004;69:158–65.
[10] TenHuisen KS, Martin RI, Klimkiewicz M, Brown PW. Formation
and properties of a synthetic bone composite: hydroxyapatite–
collagen. J Biomed Mater Res 1995;29(7):803–10.
[11] Lickorish D, Ramshaw John AM, Werkmeister JA, Glattauer V,
Howlett CR. Collagen–hydroxyapatite composite prepared by bio-
mimetic process. J Biomed Mater Res A 2004;68(1):19–27.
[12] Wang YJ, Yang CR, Chen XF, Zhao NR. Biomimetic formation
of hydroxyapatite/collagen matrix composite. Adv Eng Mater
2006;8(1–2):97–100.
[13] Brown RA, Wiseman M, Chuo CB, Cheema U, Nazhat SN.
Ultrarapid engineering of biomimetic materials and tissues: fabrica-
tion of nano- and microstructures by plastic compression. Adv Funct
Mater 2005;15(11):1762–70.
[14] Jian T, Saltzmana WM. Biomaterials with hierarchically defined
micro- and nanoscale structure. Biomaterials 2004;25:3593–601.
[15] Babini GN, Tampieri A. Towards biologically inspired materials. Br
Ceram Trans 2004;103(3):101–9.
[16] Guo XD, Yuan Q, Zheng QX, Duan ZX, Pan ZQ, Zhao M, et al.
Biologically inspired self-assembling synthesis of bone-like nano-
hydroxyapatite/PLGA-(PEG-ASP)n composite: a new biomimetic
bone tissue engineering scaffold material. J Wuhan Univ Technol,
Mater Sci Ed 2005;20(Supplement):234–7.
[17] Ryu HS, Youn HJ, Hong KS, Chang BS, Lee CK, Chung SS. An
improvement in sintering property of b-tricalcium phosphate by
addition of calcium pyrophosphate. Biomaterials 2002;23:909–14.
[18] Guo XM, Wang CY, Zhang YF, Xia RY, Hu M, Duan CM, et al.
Repair of large articular cartilage defects with implants of autologous
904 F. Zhang et al. / Acta Biomaterialia 3 (2007) 896–904
mesenchymal stem cells seeded into b-tricalcium phosphate in a sheep
model. Tissue Eng 2004;10(11–12):1818–29.
[19] Hench LL. Bioceramics. J Am Ceram Soc 1998;81:1705–28.
[20] Nagaraja S. Microstructural stresses and strains associated with
trabecular bone microdamage. PhD thesis. Georgia Institute of
Technology. December 2006, p. 139.
[21] Ignatius AA, Wolf S, Augat P, Claes LE. Composites made of rapidly
resorbable ceramics and poly(lactide) show adequate mechanical
properties for use as bone substitute materials. J Biomed Mater Res
2001;57:126–31.
[22] Koerten HK, Meulen J. Degradation of calcium phosphate ceramics.
J Biomed Mater Res 1999;44(1):78–86.
[23] Jiang GW, Shi DL. Coating of hydroxyapatite on highly porous Al2O3
substrate for bone substitutes. J Biomed Mater Res 1998;43:77–81.
[24] Deville S, Saiz E, Tomsia AP. Freeze casting of hydroxyapatite
scaffolds for bone tissue engineering. Biomaterials 2006;27:5480–9.
[25] Kaito T, Mukai Y, Nishikawa M, Ando W, Yoshikawa H, Myoui A.
Dual hydroxyapatite composite with porous and solid parts: exper-
imental study using canine lumbar interbody fusion model. J Biomed
Mater Res B 2006;78:378–84.
[26] Maciel A, Boulic R, Thalmann D. Deformable tissue parameterized
by properties of real biological tissue. In: Nicholas A, Herve D,
editors. International symposium on surgery simulation and soft tissue
modeling. Juan-les-Pins, France: Springer-Verlag; 2003. p. 74–87.