Fig 1.
Schematic illustration of (a) particle stabilized emulsions where weakly hydrophobic particles (black) stabilize oil droplets (gold) in an aqueous (blue) suspension of particles and (b) capillary suspensions where strongly hydrophobic particles produce a bicontinuous interconnected two-phase oil-water system, where perhaps the particles partially or completely transfer into the oil phase. (Note this description of the capillary suspensions is slightly different from our previous hypothesis published as Fig 1 in reference [39]).
Fig 2.
Schematic illustration of processing procedures, presenting (a) addition of polyvinyl alcohol (PVA) to water, (b) addition of ceramic powders to the PVA solution, (c) addition of surfactant cetyltrimethylammonium bromide (CTAB) to TCP suspensions, (d) thin slip casting, (f) drying and (g) lastly sintering of specimens. HA-D undergo steps a, b, e, f and g; TCP-D undergo steps a, b, c, e, f and g; TCP-E and TCP-CS undergo all steps a to g. (h) Procedure for the preparation of PCL specimens by dip-casting, showing the dipping of a stainless steel rod into a PCL melt pool, cooling and trimming of specimens to size. Colours and dimensions used in this schematic are not representative and are only for illustrative purposes.
Fig 3.
Means of percent of sample volume occupied by porosity in each formulation based on the Archimedes method, with error bars showing 95% confidence interval.
Open porosity refers to pores which are accessible externally, whereas closed porosity refers to pores which are fully within the material without pore necks to the exterior. For total and open porosities, TCP-E and TCP-CS are statistically significantly higher than TCP-D and HA-D as indicated by the asterisks (p < 0.001). p > 0.05 for all closed porosities. (Total porosity in this figure = 100%—Archimedes bulk density %TD in Table 1).
Table 1.
Density data measured by the Archimedes and geometric methods.
TD: theoretical full density.
Fig 4.
(a-f) Scanning electron micrographs of the as-fired surfaces of the β-TCP formulations at two levels of magnification. (a and d) TCP-E (emulsion) displays mainly spherical pores (highlighted by white arrows), whereas (b and e) TCP-CS (capillary suspension) has an uneven surface comprising elongated channels (highlighted by white arrows) and smaller pores. (c and f) The surface of TCP-D (dense) is generally flat with largely submicron-sized pores. (g, h, i) Confocal microscopic images with osteoblasts identified by the nuclei (blue), F-actin cytoskeleton (red) and the collagen type I (green) extracellular matrix.
Fig 5.
(a) Box plots of pore diameters–box indicates the 25% to 75% confidence interval, with the median line in the middle; whiskers show the 1.5 interquartile range; and dots beyond the whiskers denote outliers. (b-d) Histograms showing distribution of pore diameters of (b) TCP-E, (c) TCP-CS, and (d) TCP-D (with inset at a different scale). Games-Howell post-hoc test revealed that all three samples are statistically significantly different from each other with p < 0.001 for all pairwise comparisons.
Fig 6.
Graph of metabolic activity of human osteoblasts after 7 days of incubation on scaffolds of varying microstructure (TCP-E, TCP-CS and TCP-D) as well as different materials without templated porosity (TCP-D, HA-D and PCL-D).
Solid bars illustrate the means and error bars represent the standard error. All pairs show a statistically significant difference, except for TCP-E and TCP-D, TCP-E and HA-D, as well as TCP-D and HA-D.
Fig 7.
Confocal microscopy of osteoblasts on the surfaces after 7 days of incubation.
(a) TCP-E, (b) TCP-CS, (c) TCP-D, (d) HA-D and (e) PCL-D. Blue represents the cell nucleus; red dyes the F-actin cytoskeleton of the cell; green indicates the presence of Collagen Type I which is extracellular matrix laid down by viable cells.