Fabrication of porous beta-tricalcium phosphate with microchannel and customized geometry based on gel-casting and rapid prototyping.

Abstract

The tissue engineering scaffolds with three-dimensional porous structure are regarded to be beneficial to facilitate a sufficient supply of nutrients and enable cell ingrowth in bone reconstruction. However, the pores in scaffolds tend to be blocked by the cell ingrowth and result in a restraint of nutrient supply in the further side of the scaffold. An indirect approach of combining the rapid prototyping and gel-casting technique is introduced in this study to fabricate beta-tricalcium phosphate (beta-TCP) scaffolds which not only have interconnected porous structure, but also have a microchannel network inside. The scaffold was designed with customized geometry that matches the defect area, and a double-scale (micropores-microchannel) porous structure inside that is beneficial for cell ingrowth. The scaffolds fabricated have an open, uniform, and interconnected porous architecture with a pore size of 200-400 microm, and posses an internal channel network with a diameter of 600 microm. The porosity was controllable. The compressive yield strength was 4.5 MPa with a porosity of 70 per cent. X-ray diffraction analysis shows that these fabrication processes do not change the crystal structure and chemical composition of beta-TCP. With this technique, it was also possible to fabricate porous scaffolds with desired pore size, porosity, and microchannel, as well as customized geometries by other bioceramics.

Full text

Page 1

http://pih.sagepub.com/
Medicine
Engineers, Part H: Journal of Engineering in
Proceedings of the Institution of Mechanical
http://pih.sagepub.com/content/225/3/315
The online version of this article can be found at:

DOI: 10.1243/09544119JEIM769
2011 225: 315Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine
X Li, W Bian, D Li, Q Lian and Z Jin
on Gel-Casting and Rapid Prototyping
Fabrication of Porous Beta-Tricalcium Phosphate with Microchannel and Customized Geometry Based


Published by:

http://www.sagepublications.com
On behalf of:




Institution of Mechanical Engineers
can be found at:

Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine
Additional services and information for




http://pih.sagepub.com/cgi/alerts
Email Alerts:

http://pih.sagepub.com/subscriptions
Subscriptions:



http://www.sagepub.com/journalsReprints.nav
Reprints:


http://www.sagepub.com/journalsPermissions.nav

Permissions:

http://pih.sagepub.com/content/225/3/315.refs.html
Citations:
What is This?

- Mar 1, 2011 Version of Record >>
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 2

Fabrication of porous beta-tricalcium phosphate with
microchannel and customized geometry based on
gel-casting and rapid prototyping
X Li1, W Bian1, D Li1*, Q Lian1, and Z Jin1,2
1State Key Lab for Manufacturing System Engineering, Xi’an Jiaotong University, Xi’an, ShaanXi, People’s Republic of
China
2Institute of Medical and Biological Engineering, School of Mechanical Engineering, University of Leeds, Leeds, UK
The manuscript was received on 9 November 2009 and was accepted after revision for publication on 2 June 2010.
DOI: 10.1243/09544119JEIM769
Abstract:
to be beneficial to facilitate a sufficient supply of nutrients and enable cell ingrowth in bone
reconstruction. However, the pores in scaffolds tend to be blocked by the cell ingrowth and result
in a restraint of nutrient supply in the further side of the scaffold. An indirect approach of
combining the rapid prototyping and gel-casting technique is introduced in this study to fabricate
beta-tricalciumphosphate (b-TCP) scaffolds whichnot only have interconnected porous structure,
butalsohaveamicrochannelnetworkinside.Thescaffoldwasdesignedwithcustomizedgeometry
that matches the defect area, and a double-scale (micropores-microchannel) porous structure
inside that is beneficial for cell ingrowth. The scaffolds fabricated have an open, uniform, and
interconnected porous architecture with a pore size of200–400mm,and possesan internal channel
network with a diameter of 600mm. The porosity was controllable. The compressive yield strength
was 4.5MPa with a porosity of 70 per cent. X-ray diffraction analysis shows that these fabrication
processes do not change the crystal structure and chemical composition of b-TCP. With this
technique, it was also possible to fabricate porous scaffolds with desired pore size, porosity, and
microchannel, as well as customized geometries by other bioceramics.
The tissue engineering scaffolds with three-dimensional porous structure are regarded
Keywords:
scaffold, microchannel, customized geometry
1INTRODUCTION
Biodegradable scaffold in bone tissue engineering
has played an essential role of either serving as a
three-dimensional (3D) template for cell adhesion or
inducing formation of bone from the surrounding
tissue [1, 2]. The biomaterial scaffold is designed to
provide the necessary support for cells to adhere to,
proliferate, and maintain their differentiated func-
tion. Its architecture defines the ultimate shape of
the newly grown bone tissue [3, 4]. Over the past two
decades, there has been great interest in the use of
calcium phosphates, a principal inorganic constitu-
ent of natural bone, as scaffolding materials for bone
tissue engineering [5, 6]. Different phases of calcium
phosphates were utilized to fabricate porous scaf-
folds to accommodate bone tissue regeneration in
vitro or in vivo. Porous beta-tricalcium phosphate
(b-TCP), owing to its great biocompatibility, biode-
gradability, good osteoconductibility, and demon-
strated clinical efficiency, has found a wide spectrum
of applications in bone tissue engineering [7].
Previous studies have shown that the three-dimen-
sionalinterconnectedporesinscaffoldsarebeneficialto
facilitate cell infiltration, tissue regeneration, and
vascularization [8–12]. Considerable efforts have been
focused on fabricating bioceramics such as b-TCP or
HA into a three-dimensional interconnected micropor-
ous scaffold to guide bone reconstruction [13–15]. The
interconnected porous scaffolds with desired pore size
andporosityhavealreadybeenpreparedby acombina-
*Corresponding author: State Key Lab for Manufacturing System
Engineering, Xi’an Jiaotong University, Xi’an, 710049, ShaanXi,
PR China.
email: dcli@mail.xjtu.edu.cn
315
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 3

tion of gel-casting and polymer sponge methods [16].
However, these scaffolds prepared with only uniform
porous structures have difficulties providing sufficient
nutrients evenly. The nutrients, oxygen and blood,
transferred merely by infiltrating, can hardly be
delivered into the further internal areas of the scaffold,
especially the central part, and therefore result in the
death of cells in those areas. In addition, these scaffolds
with only simple geometries, mostly cylinders, restrict
their clinical applications, since it is hard for these
simple-shaped scaffolds to repair the bone defection
areas with various complex contours. With the devel-
opment of advanced manufacturing techniques, com-
puter-aided design (CAD)-based manufacturing tech-
nologies have been applied toward the fabrication of
three-dimensional scaffolds with tenable micro- and
macro-scale features [17–19]. The CAD-based solid
freeform, known as rapid prototyping (RP) techniques
that have significant advantages over traditional fabri-
cationmethods canprovide highlevelsofaccuratecon-
trol over their macro-structural (e.g. spatial form,
mechanical strength, density, porosity) and microstruc-
tural(e.g.poresize,poredistribution,poreinterconnecti-
vity) properties [20–22]. The applications of RP techno-
logies in scaffold fabrication are wide and varied; how-
ever,onlyasmallnumberhavereachedclinicaluse[23].
This study presents an approach that integrates the
gel-casting method with a RP technique which cannot
only fabricate b-TCP porous scaffolds with micro-
channel network, but can also produce customized
geometry. The microstructures of channels and pores
in scaffolds were controllable according to physiologi-
cal functions for cell infiltration, tissue regeneration,
and vascularization. Customized geometry of an
animal femur was designed based on computerized
tomography (CT) image processing and reverse en-
gineering (RE). The structural, mechanical, and che-
mical properties of the scaffolds fabricated were
studied under various microstructures. The pore
morphology, size, distribution, and the channel struc-
tures were characterized via a scanning electron
microscope (SEM). Compression tests were performed
using an Instron mechanical tester to evaluate the
yield strength and elastic modulus. X-ray diffraction
(XRD) was used to characterize the crystal structure
and chemical composition, respectively.
2MATERIALS AND METHODS
2.1Materials
The b-TCP powder, (purchased from Edward Keller
Limited, Shanghai, China), used as received, is
composed of clusters of submicron crystallites and
its particle size is in the range of 1.0 to 2.5mm with a
surface area of 1–6m2/g. The components of the gel-
casting process were reactive organic monomers:
monofunctional acrylamide, C2H3CONH2, and difunc-
tional methylenebisacrylamide, ((C2H3CONH)2CH2).
Ammonium presulfate, (NH4)2S2O8 and N,N,N9N9-
tetramethylethylenediamine (TEMED) were used as
the initiator and catalyst, respectively. A 30 per cent
aqueous solution of sodium polymethacrylate was
used as a dispersant. All these chemicals (analytic
reagent) were purchased from the Sigma-Aldrich Cor-
poration. All the slurries were aqueous, and deio-
nized(DI)waterwasusedinalltheexperiments.Paraf-
finceresinandgelatinwereutilizedforpreparingparaffin
spheres.
2.2Design
A canine was selected as the research object and one of
its knee joints was scanned using a helical CT machine.
Anatomical modelling was based on using the CT
images, Fig. 1(a), to reconstruct the 3D model of the
femur bone. Once processed, these images were
imported into Mimics Software (Materialise NV,
Belgium) and the femur skeleton was reconstructed
and visualised in a 3D display using the regional
growing technique. Point cloud data (Fig. 1(b)) repre-
sentingthefemurbone,wereobtainedbyexportingthe
3D skeleton in the IGES format. These point clouds
were imported into RE software such as Surfacer or
Geomagic, and after removing noise and reordering,
the freeform model, Fig. 1(c) of the femur bone was
reconstructed.
The negative pattern of femur scaffold was
designed with a commercial CAD software (Pro-
engineer), Fig. 2(a). The surface of the femur model
was thickened to 800mm. A central column was
designed which was 3mm in diameter running in
the vertical direction. Interconnected channels, Fig.
2(b), with a diameter of 600mm were constructed:
horizontal channels extending from the centre colu-
Table 1.
Composition of slurry for scaffold fabrication
Component Amount
Solvent:
Ceramic powder:
Monomer:
Cross linker:
Dispersant:
Initiator:
Catalyst:
Deionized water
Beta-tricalcium phosphate
Acrylamide
Methylenebisacrylamide
Sodium polymethacrylate
Ammonium presulphate
N,N,N’N’-
tetramethylethylenediamine
35g
60g
4g
0.5g
0.6g
0.2g
0.1g
316X Li, W Bian, D Li, Q Lian, and Z Jin
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.com Downloaded from

Page 4

mn towards the outside surface in radial directions,
vertical channels penetrating through the scaffold
from the top to the bottom, and the annular channels
interconnecting the horizontal channels with verti-
cal ones. The angles between the horizontal and ver-
tical channels were designed to 75u and 105u based
on conversation of energy and flow. A standard cy-
linder (diameter 20mm, height 20mm) with a similar
internal channel network was also designed for
comparison and mechanical testing.
2.3Fabrication
The moulds, designed with a microchannel network,
were fabricated on a stereolithography apparatus (SPS
600B, Xi’an Jiaotong University, Xi’an, China) with a
commercial epoxy resin (SL, 14120, Huntsman). The
CAD data of the negative pattern were converted into
STL data by Pro-engineers, imported into Rpdata
software, and converted into an input file for stereo-
lithography. The fabricated moulds were then cleaned
with isopropanol alcohol, Fig. 2(c) and Fig. 2(d).
Into a large beaker were added 60g paraffin block
and 3g gelatine with 1L DI water, which was heated
to 90uC. When all the paraffin was liquefied, the
solution was stratified to diphase, which, while being
maintained at the temperature of 80uC, was stirred
at 450r/min with a magnetic stirring apparatus for
about 30min, and then the microliquid droplets of
paraffin were distributed in the water evenly. The
stirring solution was instantly poured into 3L ice
water and the paraffin spheres were formed. After
being dried at room temperature under vacuum,
paraffin spheres with desired size were selected by
sieving. Figure 3 shows the paraffin spheres with
diameters ranging from 200 to 400mm.
Along with monomers (acrylamide, methylenebi-
sacrylamide), dispersant (sodium polymethacrylate)
b-TCP powders were mixed with DI water to form
the ceramic slurry. Table 1 shows the amount of
chemicals added to DI water to formulate the
ceramic slurry. The slurry prepared was deagglom-
erated by ultrasound for 5h and subsequently de-
aired under vacuum until there was no further
Fig. 1
Customized geometry design based on CT images using RE: (a) original CT image, (b)
point cloud of femur bone, (c) freeform of femur bone
Fig. 2
The CAD mould and resin mould with microchannels: (a) CAD model of mould, (b) CAD
model of channel network, (c) resin mould, (d) vertical section of resin mould
Fabrication of porous beta-tricalcium phosphate317
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 5

release of air bubbles from the sample. Catalyst
(ammonium presulfate) and initiator (N,N,N9N9-
tetramethylethylenediamine) were added to the
slurry to polymerize the monomers. The amount of
the catalyst and initiator was controlled to allow a
sufficient time for the casting process.
The paraffin spheres were filled into the moulds
and subsequently pre-adhered in a drying oven at a
temperature of 50uC for 1h. Then the b-TCP slurry
was cast into the moulds under vacuum to force the
b-TCP powders to migrate into the interspaces of the
paraffin spheres. Figure 4(a) shows the mould after
the casting process. The samples were dried at room
temperature for 72h. After the drying, pyrolysis of
the epoxy resin moulds and paraffin spheres was
conducted in air in an electric furnace with a heating
rate of 5uC/h from room temperature to 340uC,
holding 5h at 340uC to ensure most paraffin spheres
were burned out, and then sintered to 660uC at a
rate of 10uC/h, holding 5h at 660uC to ensure most
of the epoxy resin was burned out. After that the
heating rate went up to 60uC/h up to 1200uC,
holding 5h at 1200uC, and then decreased to room
temperature in 48h. Fig. 4(b) shows the scaffold
after sintering.
2.4Characterization
The density measurements provided information
about pore size and distribution, permeability, and
presence of structural faults in sintered ceramic
structures [24]. The porosity and density of scaffolds
weremeasuredwiththe
method. A scaffold of weight W was immersed in a
graduated cylinder containing a known volume (V1)
of ethanol. The cylinder was placed in a vacuum to
force the ethanol into the pores of the scaffold until
no air bubbles emerged from the scaffold. The total
volume of the ethanol and scaffold was then
recorded as V2. The volume difference (V22V1)
was the volume of the skeleton of the scaffold. The
scaffold was removed from the ethanol and the
residual ethanol volume was measured as V3. The
total volume of the scaffold was (V22V3). The
apparent density of the scaffold was evaluated as W/
(V22V3). The porosity of the open pores in the
scaffold was evaluated as (V12V3)/(V22V3).
A SEM (S-3000, Hitachi, Japan) was used for
morphological characterization of the scaffolds.
The samples were coated with gold/palladium under
an argon atmosphere and the magnification times
were set from 50 to 300. An Instron 5848 (Canton,
MA, USA) mechanical tester with 10kN load cell was
used for the compression mechanical test using the
guidelines set in ASTM. The crosshead speed was set
at 0.500mm/min, and the load was applied until the
scaffold cracked. The yield strength was determined
from the crosspoint of the two tangents on the
stress–strain curve around the yield point. Thermo-
gravimetric analysis (TGA) was used to measure
thermal stability and composition of a material. It
measures weight changes in a material as a function
of temperature (or time) under a controlled atmo-
sphere. TGA was used to study the pyrolysis process
of epoxy resin and paraffin and determine the
temperature of burning them off. It was performed
in a vertical tube furnace, with a heating rate of
2.5uC/min up to 800uC under nitrogen flow. XRD
was used to characterize the crystallinity, chemical
liquiddisplacement
Fig. 3
The micrograph of paraffin spheres
Fig. 4
Photograph of fabricating process: (a) mould
filled with slurry after gel-casting, (b) scaffold
after sintering
318X Li, W Bian, D Li, Q Lian, and Z Jin
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 6

composition, and structure of the materials. XRD
experiments were performed on as-received b-TCP
powders, before-sintered samples, and sintered sam-
ples after being crushed to powders, using CuKa
radiation at 20mA, 40kV. Scans were performed
between 2h values of 10u and 70u at a rate of 0.4u/min.
3 RESULTS AND DISCUSSIONS
An indirect approach of combining the rapid proto-
typing and gel-casting is introduced in this study,
which provides a better control over the complex
geometries and microstructures. The process of
scaffold fabrication is outlined in Fig. 5. Scaffolds
with different porosities (30, 50, 70 per cent) and
different types (with channel network inside, with-
out channel network inside) were prepared to
evaluate the effect of porosity and channel network
on physical and mechanical properties. High poros-
ity is required for better cell ingrowth while the
mechanical performance is poor.
Figure 6 shows that the yield strength of the scaffold
subsides from ,18MPa with 30 per cent porosity to
,4.5MPa with 70 per cent porosity, which corre-
sponds to the mechanical behaviour of human
cancellous bone. The yield strength of human cancel-
lous bone is reported to differwithin the range of 0.6 to
17.5MPa [25, 26]. The mechanical properties of a
porous material depend on the density of the porewall
material as suggested by Gibson [27, 28]. The porosity
is higher, the mechanical behaviour is weaker. The
porosity and mechanical properties should be ba-
lanced in future work by the current authors.
Figure 7 shows that the apparent density of scaffold
decreases from ,1.7g/cm3with 30 per cent porosity
to ,0.7g/cm3with 70 per cent porosity. The channel
network has slight effects on both yield strength and
apparent density. The apparent density of a porous
scaffold can influence its mechanical strength, perme-
ability, and presence of structural defects [29]. The
resin mould with a desired geometry and microchan-
nels was filled with paraffin spheres, which were
adopted to produce interconnected pores.
Figure 8 shows the adhering process of paraffin
spheres. The inter-diameter can be controlled by
Fig. 5
A flow chart of process steps for scaffold
fabrication using combined gel-casting and
rapid prototyping
Fig. 6
Yield strength of b-TCP scaffolds as a function of the porosity (Sample: 3)
Fabrication of porous beta-tricalcium phosphate 319
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013 pih.sagepub.comDownloaded from

Page 7

adjusting the temperature and pressure during the
adhering process. The b-TCP slurry was cast into the
moulds under vacuum. Then, both the moulds and
paraffin spheres were removed by pyrolysis. The
scaffolds fabricated not only have open, uniform,
and interconnected porous architectures, but also
posses an internal channel network. It is reported
that the pore size with a diameter of 100–400mm is
beneficial for cell infiltration and host tissue in-
growth [30, 31]. Considering if the pore size is too
small, the pores could be blocked with the cell
ingrowth. Therefore the double-scale porous struc-
ture with a pore diameter of 200–400mm and
channel diameter of 600mm was adopted. The size
and structure of microchannel should be optimized
in future work by the current authors. The mico-
channel network can be seen in Fig. 9(a), the vertical
section of the scaffold. Figure 9(b) shows the optical
photomicrograph of the scaffold. The morphology of
the interconnected pores can be seen in Figs 9(c)
and (d), the SEM micrographs of porous scaffold.
Pyrolysis of the resin as well as the paraffin is
critical in sintering ceramics by the indirect techni-
que. Sufficient time should be given for the resin and
Fig. 9
The photographs and micrographs of scaffold:
(a) vertical section of the scaffold, (b) optical
photomicrograph of the scaffold, (c, d) SEM
micrograph of pores in scaffold
Fig. 7
Apparent density of b-TCP scaffolds as a function of the porosity (Sample: 3)
Fig. 8
The adhering process of paraffin spheres
320X Li, W Bian, D Li, Q Lian, and Z Jin
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013 pih.sagepub.comDownloaded from

Page 8

paraffin burn out before the sintering of ceramics
starts to avoid cracks in the microstructure. In this
study, TGA was used to determine the temperature
at which the complete burnout of the resin and
paraffinoccurred. Figure 10 shows the weight
change of the resin and paraffin with temperature.
It is seen that the resin and paraffin were completely
burned out at 660uC and 340uC respectively. Thus,
to allow ample time for the complete burnout of the
resin and paraffin in scaffolds before the sintering
started, and to ensure no cracking occurs, the
heating process was set to 5uC/h from room
temperature to 340uC with a dwell time of 5h, then
sintered to 660uC at a rate of 10uC/h with a dwell
time of 5h. After that the heating rate went up to
60uC/h up to 1200uC with a dwell time of 5h, and
then decreased to room temperature in 48h. Al-
though it has been demonstrated that pure b-TCP
is biocompatible and nontoxic, variations in the precise
nature of the calcium phosphate phases can have a
strong effect on the cellular response and thus reduce
the material’s biocompatibility [32, 33]. Moreover,
changes in the degree of crystallinity and phase
purity may also lead to variations in the level of spe-
cimen solubility, which would affect the scaffold
degradation [33–35]. Therefore, preserving the phase
composition and crystalline structure of the b-TCP
during sintering becomes critical for its biological
applications.
Figure 11 shows the XRD patterns of the b-TCP
powder as received, before sintering, and after sinter-
ing at 1200uC. The XRD peaks of before-sintered and
after-sintered diffraction patterns agree well with the
as-received pattern. No discernible difference was
observed, and no additional phase was identified. This
indicates thatthe casting and sintering processhas not
changed the composition of b-TCP. To validate the
biocompatibility of the material and to optimize the
porous structure and network on cell adhesion,
proliferation, and growth, the bio-test both in vitro
and in vivo will be part of the current authors’ future
work.
4 CONCLUSIONS
This study has developed a method for fabricating
porous scaffold with microchannel network and
customized geometry by combining gel-casting and
rapid prototyping techniques. A yield strength of
,4.5MPa for the scaffold with a porosity of 70 per
Fig. 10
Weight loss as a function of temperature for
pyrolysis of resin and paraffin under a nitro-
gen environment
Fig. 11
XRD patterns for: (a) b-TCP powder as received, (b) scaffold before sintered, (c) scaffold
Fabrication of porous beta-tricalcium phosphate321
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 9

cent was achieved. The porous scaffold, with a pore
size ranging from 200 to 400mm, not only has
customized geometry of a dog femur, but also
possesses an internal channel network with a
diameter of 600mm. Because the microchannels
and geometry were produced based on RP techni-
que, the desired structure of the channel network
and complex shape can be fabricated. The pore size,
shape, and porosity were controllable since the
porous structures were created by paraffin spheres.
The inter-diameter can be controlled during the
adhering process of paraffin spheres. XRD results
showed that the phase, crystallinity of the b-TCP
scaffolds remain unchanged after sintering. With this
technique, other bioceramics can be applied to
fabricate tissue engineering scaffolds with controlled
microstructures and customized shapes.
ACKNOWLEDGEMENTS
This work was supported by grants from the Natural
Science Foundation of China (50628505, 50775178),
Natural Science Foundation of Shaanxi Province
(SJ08E110), and by the Program for Changjiang
Scholars and Innovative Research Team in Univer-
sity (IRT0646).
F Authors 2011
REFERENCES
1 Vacanti, C. A. and Vacanti, J. P. Bone and cartilage
reconstruction with tissue engineering approaches.
Otolaryngologic Clin. North Am., 1994, 27(1), 263–
276.
2 Vacanti, C. A. and Bonassar, L. J. An overview of
tissue engineered bone. Clin. Orthop. Related Res.,
1999, 367, S375–S381.
3 Liu, X. H. and Ma, P. X. Polymeric scaffolds for
bone tissue engineering. Annls Biomed. Engng,
2004, 32(3), 477–486.
4 Hutmacher, D. W. Scaffold design and fabrication
technologies for engineering tissues – state of the
art and future perspectives. J. Biomater. Sci. –
Polymer Edition, 2001, 12(1), 107–124.
5 LeGeros, R. Z. Properties of osteoconductive bioma-
terials: calcium phosphates. Clin. Orthop. Related
Res., 2002, 395, 81–98.
6 Bucholz, R. W., Carlton, A., and Holmes, R. E.
Hydroxyapatite and tricalcium phosphate bone-
graft substitutes. Orthop. Clin. North Am., 1987,
18(2), 323–334.
7 Ogose, A., et al. Histological examination of beta-
tricalciumphosphate
J. Biomedical Mater. Res., 2002, 63(5), 601–604.
graftinhuman femur.
8 Zhang, Z., et al. Pore size, tissue ingrowth, and
endothelialization of small-diameter microporous
polyurethane vascular prostheses. Biomaterials,
2004, 25(1), 177–187.
9 Zeltinger, J., et al. Effect of pore size and void
fraction on cellular adhesion, proliferation, and
matrix deposition. Tissue Engng, 2001, 7(5), 557–
572.
10 O’Brien, F. J., et al. The effect of pore size on cell
adhesion in collagen-GAG scaffolds. Biomaterials,
2005, 26(4), 433–441.
11 Hutmacher, D. W., Sittinger, M., and Risbud, M. V.
Scaffold-based tissue engineering: rationale for com-
puter-aided design and solid free-form fabrication
systems. Trends in Biotechnol., 2004, 22(7), 354–362.
12 Chen, C. S., Mrksich, M., Huang, S., Whitesides, G.
M., and Ingber, D. E. Geometric control of cell life
and death. Science, 1997, 276(5317), 1425–1428.
13 Yoshikawa, T., Ohgushi, H., and Tamai, S. Im-
mediate bone forming capability of prefabricated
osteogenic hydroxyapatite. J. Biomedical Mater.
Res., 1996, 32(3), 481–492.
14 Ohgushi, H., Okumura, M., Yoshikawa, T., Inoue,
K., Senpuku, N., Tamai, S., et al. Bone-formation
process in porous calcium-carbonate and hydro-
xyapatite. J. Biomedical Mater. Res., 1992, 26(7),
885–895.
15 Li, S. H., De Wijn, J. R., Layrolle, P., and de Groot,
K. Synthesis of macroporous hydroxyapatite scaf-
folds for bone tissue engineering. J. Biomedical
Mater. Res., 2002, 61(1), 109–120.
16 Ramay, H. R. and Zhang, M. Q. Preparation of
porous hydroxyapatite scaffolds by combination of
the gel-casting and polymer sponge methods.
Biomaterials, 2003, 24(19), 3293–3302.
17 Tsang, V. L. and Bhatia, S. N. Three-dimensional
tissue fabrication. Advd Drug Delivery Rev., 2004,
56(11), 1635–1647.
18 Sachlos, E., Reis, N., Ainsley, C., Derby, B., and
Czernuszka, J. T. Novel collagen scaffolds with
predefined internal morphology made by solid free-
formfabrication.Biomaterials,2003,24(8),1487–1497.
19 Lam, C. X. F., Mo, X. M., Teoh, S. H., and
Hutmacher, D. W. Scaffold development using
3D printing with a starch-based polymer. Mater.
Sci. Engng C – Biomimetic and Supramolecular
Systems, 2002, 20(1–2), 49–56.
20 Wilson, C. E., de Bruijn, J. D., van Blitterswijk,
C. A., Verbout, A. J., and Dhert, W. J. A. Design and
fabrication of standardized hydroxyapatite scaf-
folds with a defined macro-architecture by rapid
prototyping for bone-tissue-engineering research.
J. Biomedical Mater. Res. Part A, 2004, 68A(1),
123–132.
21 Singare, S., Dichen, L., Bingheng, L., Zhenyu, G.,
and Yaxiong, L. Customized design and manufac-
turing of chin implant based on rapid prototyping.
Rapid Prototyping J., 2005, 11(2), 113–118.
22 Leong, K. F., Cheah, C. M., and Chua, C. K. Solid
freeform fabrication of three-dimensional scaffolds
322X Li, W Bian, D Li, Q Lian, and Z Jin
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from

Page 10

for engineering replacement tissues and organs.
Biomaterials, 2003, 24(13), 2363–2378.
23 Hutmacher, D. W. and Cool, S. Concepts of
scaffold-based tissue engineering-the rationale to
use solid free-form fabrication techniques. J.
Cellular and Molecular Med., 2007, 11(4), 654–669.
24 Hodgskinson, R. C. J. The effect of variation in
structure on the Young’s modulus of cancellous
bone: a comparison of human and non-human
material. Proc. IMechE, Part H: J. Engineering in
Medicine, 1990, 204(2), 7.
25 Kopperdahl, D. L. and Keaveny, T. M. Yield strain
behavior of trabecular bone. J. Biomechanics, 1998,
31(7), 601–608.
26 Morgan, E. F. and Keaveny, T. M. Dependence of
yield strain of human trabecular bone on anatomic
site. J. Biomechanics, 2001, 34(5), 569–577.
27 Gibson, L. J. Biomechanics of cellular solids.
J. Biomechanics, 2005, 38(3), 377–399.
28 Gibson, L. J. Cellular solids. Mrs Bulletin, 2003,
28(4), 270–271.
29 Sepulveda, P., Ortega, F. S., Innocentini, M. D. M.,
and Pandolfelli, V. C. Properties of highly porous
hydroxyapatite obtained by the gelcasting of foams.
J. Am. Ceramic Soc., 2000, 83(12), 3021–3024.
30 Yamamoto, M., Tabata, Y., Kawasaki, H., and
Ikada,Y.Promotion offibrovascular tissueingrowth
into porous sponges by basic fibroblast growth
factor. J. Mater. Sci. Mater. Med., 2000, 11(4),
213–218.
31 Hulbert,S.F.,Morrison,S.J.,andKlawitte,J.J.Tissue
reaction to 3 ceramics of porous and non-porous
structures. J. Biomedical Mater. Res., 1972, 6(5),
347–374.
32 Best, S., Sim, B., Kayser, M., and Downes, S. The
dependence of osteoblastic response on variations
in the chemical composition and physical proper-
ties of hydroxyapatite. J. Mater. Sci. Mater. Med.,
1997, 8(2), 97–103.
33 Eggli, P. S., Muller, M., and Schenk, R. K. Porous
hydroxyapatite and tricalcium phosphate cylinders
with 2 different pore-size ranges implanted in the
cancellous bone of rabbits – a comparative histo-
morphometric and histologic-study of bony in-
growth and implant substitution. Clin. Orthop.
Related Res., 1988, 232, 127–138.
34 Dhert, W. J. A., Klein, C. P. A. T., Wolke, J. G. C.,
Vandervelde, E. A., Degroot, K., and Rozing, P. M.
A mechanical investigation of fluorapatite, magne-
siumwhitlockite,and
sprayed coatings in goats. J. Biomedical Mater.
Res., 1991, 25(10), 1183–1200.
35 Holmes, R., Mooney, V., Bucholz, R., and Tencer,
A. A Coralline hydroxyapatite bone-graft substitute
– preliminary report. Clin. Orthop. Related Res.,
1984 (188), 252–262.
hydroxylapatiteplasma-
Fabrication of porous beta-tricalcium phosphate323
Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine
at ETH Zurich on June 4, 2013pih.sagepub.comDownloaded from