Effect of hydroxyapatite on biodegradable scaffolds

fabricated by SLS
Szilvia Eosoly 1, Stefan Lohfeld 2and Dermot Brabazon 1
1 School of Mechanical and Manufacturing Engineering, Dublin City University, Dublin 9, Ireland
2 National Centre for Biomedical Engineering Science, National University of Ireland,
Galway,Ireland

Abstract.
Selective laser sintering (SLS) has the potential to fabricate bioresorbable polymer / ceramic
composite scaffolds with pre-designed external and internal architecture that can be used for
bone tissue engineering applications. Scaffolds were fabricated using poly-c-caprolactone as
the base material. The effect of 15 and 30 wt% of hydroxyapatite (HA) addition was
investigated in terms of compressive properties, accuracy, surface topology, and wettability.
Fabricated dimensions of PCL microstructures showed great deviations from their nominal
values. Average surface roughness was found to be Ra=25±4 gm. Increased HA content had
no statistically significant effect on accuracy and surface roughness. However the addition of
HA had a significant influence on compressive properties, hydrophobicity and wettability of
the samples. Addition of 30 wt% HA improved initial compressive modulus of pure PCL
scaffolds from 1.31±0.08 MPa to 1.58±0.18 MPa. Yield strength values increased from
0.14±0.07 MPa to 0.17±0.01 MPa by adding 15 wt% of HA, but decreased with further HA
addition. Yield strain for all compositions was over c=0.06. Increased HA content decreased
hydrophobicity and increased wettability of scaffold surfaces. The study demonstrated the
ability of SLS to fabricate tissue engineering scaffolds, and the positive effect of HA particle
reinforcement in terms of compressive mechanical properties and surface characteristics.

Introduction
Selective Laser Sintering (SLS) is a promising technology for fabricating complex scaffold
geometries with pre-designed, intricate internal architectures for applications in tissue
engineering. SLS is a powder based solid freeform fabrication technology. 3D models are
created in CAD environment and sliced into thin layers. The geometry is reconstructed layer-
by-layer using a CO2 laser beam to sinter the layers according to the slice data. In the scanned
regions the sintering process is characterized by particle coalescence resulting from the
delivered heat and energy. As a result of particle coalescence, reduction in surface free
energy is the main driving force for the sintering process [1].
Bone tissue engineering scaffolds aim to temporarily replace the functions of the extracellular
matrix of natural bone, to which cells can attach, proliferate and form tissues. Mechanical and
microstructural characteristics and surface morphology have been shown to strongly
influence cell adhesion, growth and differentiation [2-3]
Several papers reported successful selective laser sintering of bone scaffolds using
biodegradable polymer/ceramic composites [4-5], however quantitative characterization of
accuracy, surface and mechanical properties is less investigated. In this study poly-c-
caprolactone (PCL) and hydroxyapatite (HA) were selected for selective laser sintering hard
tissue scaffolds. Accuracy, surface morphology and mechanical properties of the fabricated
PCL scaffolds are examined as a function of increasing HA content. Addition of HA is
expected to enhance surface morphology including bioactivity and wettability of the scaffolds
and to improve their mechanical behavior.

Materials and Methods

Materials. Poly-c-caprolactone (PCL, Sigma Aldrich Chemical Co.) pellets were
cryogenically grinded and sieved. The used powder batch had an average particle size of 125
m with particle size distribution where 80 % of all particles are between 70-160 m as
measured by Malvern Mastersizer particle size analyzer. The hydroxyapatite (HA) powder
used in this experiment is sold under brand name Captal 60-1 (Plasma Biotal Ltd.) and has an
average particle size of 38 gm. Mixtures of PCL/HA powders were produced by physical
blending. PCL was used as the base material and HA was added to produce blends with 0, 15
and 30 wt% HA content.
Geometry. Test specimens for mechanical testing, surface roughness and contact angle
measurements were designed to be circular solid discs with diameter and thickness of 12 mm
and 3.5 mm, respectively. Additionally, for mechanical testing and accuracy measurements
lattice structures with a relative density of 0.33, strut size of 0.6 mm and pore size of 1.2 mm
were created. All specimens were designed using SolidWorks and were exported to STL file
format.
Scaffold Fabrication. All specimens were fabricated on a DTM Sinterstation 2500 Plus (3D
Systems) using the same process parameter settings - layer thickness of 0.15 mm, part bed
temperature of 35 °C, laser fill power of 10 W, outline laser power of 5 W and scan spacing
of 0.15 mm.
Accuracy. Dimensions of struts manufactured in different building directions were obtained
from scanning electron micrographs (Zeiss EVO LS15).
Surface characteristics. Surface topology and accuracy of the fabricated specimens were
examined using scanning electron microscopy (SEM). Surface profile was obtained using
optoNCDT laser profiler. Obtained data was analyzed in Matlab7.1 and surface roughness
(Ra) was calculated.
Mechanical properties. Compressive mechanical properties of the fabricated specimens
were tested to ISO 604 standards using a 5 kN load cell at crosshead speed of 1 mm/s and
applying 1 N preload. Compressive modulus, yield strength and yield strain values were
reported. Scaffolds were tested in a way that loading direction was parallel to struts fabricated
in the X-building direction.
Contact Angle. The contact angle of water on the sintered surfaces was measured using
ArtCAM 130 MI BW monochrome camera, and FTA200 contact angle analyzer software.
Static approach was used to analyze initial contact angle, and contact angle evolution after 1,
2 and 3 s.
Statistics. All measurements were performed on 7 replicates. Result data was expressed as
mean±standard deviation. Comparisons between groups were made using one-way ANOVA
analysis. Statistical significance was set at P<0.05.

Results and discussions

Accuracy. The struts within the scaffolds showed great dimensional variance in the three
manufacturing directions. Struts built in the X, Y and Z direction were measured to be
582±242, 977±36, 730±101 pm, respectively. High standard deviation values are the results
of manufacturing induced deformations. All struts were designed to have a square cross
section of 600 x 600 m, however, the cross section of a single, randomly selected strut built
in the X direction was reproduced as 411 x 750 pm rectangle with fillets on the corner. This
is the result of the different scanning patterns in different building directions. Poor accuracy
of micro-features fabricated by commercial SLS machines is also due to the fact the laser
beam diameter (0.4 mm) is in the range of the fabricated feature sizes (0.6 mm).
The addition of HA had no statistically significant effect on the accuracy of the samples,
however standard deviations increased with higher HA content, indicating decrease in
reproducibility.

Surface properties. It has been shown that rough surfaces and high surface to volume ratio is
favorable for cell attachment [2]. In terms of tissue engineering, an advantage of powder
based manufacturing technologies including SLS, is that fabricated part surfaces are rough.
Surface roughness is dependent on the particle size of the processed powder. Surface
roughness (Ra) of PCL scaffold was 25±4 inn and was not significantly altered by the
addition of HA. SEM images (Fig. 1, Fig. 2a) show neck formation between PCL particles
indicating that the final densification stage of sintering where pores are elmininated from the
designed solid regions was not reached. Therefore accuracy and roughness of the scaffolds
were mainly determined by the particle size (125μm) of the PCL that was greater than that of
HA (38 μm).
Contact angle. Initial contact angle of pure PCL scaffolds was 104±5.5° and was constantly
decreased by the addition of HA. The initial contact angle for composites with 15 wt% of HA
decreased to 98±13.6° and with 30 wt% HA to 83±16.9°. Results demonstrated the ability of
added HA powder to reduce hydrophobicity of the PCL scaffolds. Contact angle of water on
the surface of parts with 0 and 15 wt% HA did not change significantly in the examined 3 s
interval; however specimens with 30 wt% of HA were completely wettable as the water drop
was soaked in within 3 s. The hydrophobicity of the scaffolds was decreased and wettability
was increased by the addition of HA making them more attractive for cell attachment [3].

Mechanical properties. It has to be noted that microporosity was present both in the
designed solid disks and in the designed solid regions of the sintered scaffolds. Average
measured relative density of the disks was 0.47. For the scaffolds the overall measured and
designed values were 0.3 and 0.33, respectively, indicating that microporosity of the struts is
lower than that of the disks. Reduction in porosity within the designed solid regions for parts
with smaller feature sizes is the result of higher energy density being delivered to a given
surface area due the more frequent outline scans.

For the designed solid disks increase in HA content resulted in a linear increase in
compressive modulus from 3.75±0.38 MPa for pure PCL scaffolds to 5.58±0.62 MPa for
scaffolds with 30 wt% of HA content. Although in particle reinforced composites mainly the
matrix bears the loading, HA particles inhibit the movement of molecular chains in the matrix
[6], explaining the observed improvement of mechanical properties.
However, HA particles would agglomerate at higher concentrations, resulting in deterioration
of compressive properties. In the compressive behavior of fabricated scaffolds three distinct
regions can be identified [7]. As shown in Fig. 2b, the first region was a linear elastic slope
characterized by elastic compression of the struts, followed by a plastic collapse region that
marks the buckling of struts and formation of plastic hinges, finally further stress compressed
the solid material itself, resulting in steep densification (Fig 2).
Compressive properties for the scaffolds are summarized in Table 1. Addition of 15 wt% HA
significantly improved their mechanical properties. With addition of 30 wt% HA yield
strength decreased, however it must be noted that yield strain for all materials was over 6 %
that is significantly greater than typical strains during physiological loading. Nevertheless the
observed compressive properties are not sufficient for load bearing bone scaffolds but can be
further improved by post processing.

Conclusions
The study demonstrated that bone scaffolds with predesigned external and internal
architectures can be manufactured using SLS. The addition of HA made the PCL scaffold
more suitable for tissue engineering applications by increasing mechanical properties and
decreasing hydrophobicity. Compressive properties of scaffolds should be further improved
by post processing of sintered samples.

Acknowledgments
This research has been supported by a Marie Curie Early Stage Research Training
Fellowship of the European Community's Sixth Framework Programme under contract
number MEST-CT-2005- 020621.

References
[1] R.M. German: Sintering theory and practice (Wiley-Interscience Publication, USA 1996)
[2] B. D. Boyan, T. W. Hummert, D.D Dean, Z. Schwartz: Biomaterials Vol. 17 (1996)
p.137-146 [3] K.Webb, V. Hlady, P.A. Tresco: Biomed. Mater. Res Vol.41 (1998), p.422-430
[4] F.E. Wiria, K.F. Leong, C.K. Chua, Y. Liu: Acta Biomaterialia Vol.3 (2007) p.1-12
[5] K.H. Tan, C.K. Chua, K.F. Leong, C.M. Cheah, P. Cheang, M.S. Abu Bakar, S.W.Cha:
Biomaterials Vol 24 (2003) p.3115-3123 [
6] B. Chen, K. Sun: Polymer Testing Vol.24 (2005) p.978-982
[7] L.J. Gibson, M.F. Ashby: Cellular solids: structure and properties (Cambridge University
Press, UK 1997)