Page 74 - IJB-1-1

A novel bioactive PEEK/HA composite with controlled 3D interconnected HA network

respectively. As seen in Figure 3(B), pores within the result in inadequate melting of PEEK and therefore,
range of 1–5 µm in both fractured surface and external failure to perfuse the scaffold pores at low pressure.
surfaces of the filament were realized through sinter- Dwelling time must be sufficient to permit adequate
ing. The rough surface of HA filaments has the poten- heat transfer throughout the powder, including powder
tial to promote osteointegration in vivo. The main residing at a distance from the heat source. Conversely,
advantage of solvent-based extrusion freeforming is prolonged time at the target temperature will result in
that both macroporosity of scaffold or spacing (yellow thermal oxidation and degradation of the polymer.
areas in Figure 3(A)) and microporosity of filaments Samples with 400 µm filament and 500 µm pore sizes
(in Figure 3(B)) are controllable. Macroporosity was were overmoulded using dwelling times of 4, 12, 16,
controlled via computer design, and microporosity via and 20 minutes, under dynamic loading (details in
alternation in sintering temperature and dwelling time. Table 1). Four minutes dwelling time was insufficient
to melt the PEEK, 12 minutes resulted only in partial
melting (Figure 4(A)), and 16 minutes dwelling time
b
resulted in complete melting of the PEEK, but
inadequate scaffold infiltration (Figure 4(B)). It was

determined that 20 minutes was the optimal dwelling
time for compression moulding of PEEK at 400℃,
permitting adequate melt flowability to fill the lattice
structures with no apparent polymer degradation.
PEEK degradation was realized by visual observation
Figure 3. (A) A typical sintered HA scaffold with uniform mi- (i.e., colour change in PEEK) in this study.
crostructure and macroporosity (yellow rectangles), (B) a mag-
nified image of the selected region (red rectangle) which in-
cludes both external and internal surface of filaments.

Like most conventional polymers, the viscosity of
PEEK decreases with increasing temperature. The
melting temperature of PEEK is approximately 340℃,
and at temperatures of 360℃ to 400℃, the shear vis-
3
3
cosity of PEEK varies from ~77×10 Pa∙s to ~66×10
Pa∙s. Therefore, to enable minimum shear viscosity, a Figure 4. Bioactive PEEK/HA composites, scaffolds size 10 ×
mould temperature of 400℃ was used. Conrad et 10 × 3 mm, filament size: 400 µm, pore size 500 µm, moulding
al. [26] demonstrated that this increased mould temper- temperature: 400ºC, pressure: 0.39 MPa, dynamic loading for 5
ature may also increase the compressive modulus, s; (A) dwelling time: 12 min, heating rate: 20℃/min, and (B)
yield strength and strain of the traditional PEEK/HA dwelling time: 16 min, heating rate: 20℃/min.
composites. Working temperatures greater than 380℃ 3D printed HA scaffolds are relatively fragile and
have been shown to reduce crystallinity, which is moulding pressure must be carefully regulated to per-
known to be beneficial for ductility and mit flow of molten PEEK and perfusion of fine
toughness [27,28] , with no adverse effect on pores without resulting in fracturing of the HA net-
biocompatibility in vivo [27] . Given the positive effects work. The results of our previous mechanical tests on
described on both shear viscosity and mechanical extrusion freeformed HA scaffolds indicated that

properties of the resulting PEEK samples, a mould compressive loading must be less than 1 MPa [29] ,
temperature of 400℃ was selected for this study. which is remarkably less than what has been reported
While mould temperatures of 400℃ guarantee for moulding of PEEK/HA powders. Roeder’s group
minimum viscosity, caution must be exercised when densified PEEK and HA-whisker dry powders at 125
working at temperatures in excess of 380℃ as there is MPa first, to avoid porosity, and then compression
[8]
a risk of thermal oxidation [26] . Potential of oxidation moulded at 250 MPa and 350℃–370℃ . Wong et
[5]
can be decreased by a reduction in dwelling time at al. described an alternative technique in which
the target temperature; as such optimisation of dwel- PEEK and strontium-containing HA (Sr-HA) powders
ling time is crucial. An insufficient dwelling time will were densified at 35 MPa and then compression

70 International Journal of Bioprinting (2015)–Volume 1, Issue 1