In Vitro Dissolution of Bioactive Glasses with the Focus on S53P4-Derived Products

Glass has been used for thousands of years in multiple forms and applications. Its overall importance in everyday life gives a wide variety of applications such as tableware products, windows, light bulbs, and in technical applications, e.g., telescope lenses, computer screens, and optical fibres. Glass is also used in medical applications as implants. Bioactive glass is a glass that bonds chemically with the bone and also stimulates bone growth. As the bioactive glass implant dissolves, the bone-bonding mechanism is attributed to forming a bone-like hydroxyapatite layer on the glass surface after several reaction steps. Various hot-worked products can be made of bioactive glasses, e.g., granules, microspheres, scaffolds, and fibres. Today, bioactive glasses are used mainly as granules in clinical applications. During the past years, bioactive glass microspheres have been studied as components in paste-like injectable bone grafting materials, i.e., putties, to enhance moulding properties.

This thesis investigated the properties of products made of silica-based bioactive compositions, focusing on bioactive glass S53P4 granules, microspheres, and porous scaffolds. Controlled dissolution is a crucial property of bioactive glass implants. The dissolution of bioactive glasses is a complex process influenced by various factors, including the glass composition, morphology, local environment around the implant, and surface area implant-to-volume solution ratio. In vitro dissolution is traditionally investigated by immersing the bioactive glass sample in a static, buffered solution. However, dynamic conditions, which better mimic the environment around the bioactive glass implant in the body, would provide more relevant results. Continuous flow-through experiments, where a fresh solution is continuously fed through a particle bed, allow for measuring the ion release and glass dissolution over time. In this thesis, glass microspheres with a well-defined surface area were used to study the dissolution kinetics. Static and dynamic dissolution experiments of commercial flame-sprayed S53P4-based microspheres with a size fraction of 45-500 µm were carried out in TRIS buffer and simulated body fluid for up to one week to gain a better understanding of the in vitro reactions leading to the formation of reaction layers (silica-rich and CaP). Significant differences in microspheres’ composition depending on the size fraction led to large differences in the dissolution behaviour. The smallest 45-90 µm spheres appeared mostly inert and dissolved slower than larger microspheres, while the largest spheres reacted similarly to crushed glass particles.

Microspheres with compositions close to the bioactive glasses S53P4 and 13-93 were successfully fabricated with experimental flame-spraying equipment. Under continuous fluid flow conditions, the microspheres of both compositions showed typical reaction layers, verifying their desired in vitro bioactivity. When developing new bioactive glass compositions, which combine high bioactivity and low crystallisation tendency, a deeper understanding of the role of the glass structure modifiers on the structural, thermal, and in vitro properties is required. For this, the impact of gradually replacing Na2O with K2O in the bioactive glass S53P4 was studied. Substituting 20% of K2O for Na2O led to the narrowest hot-working range, i.e., the easiest crystallisation during the processing. Replacing 66% of Na2O with K2O suggested the highest ion release and, thus, the highest pH change in static dissolution in TRIS at 24 h. However, this composition gave the lowest pH change, silica release, and the thinnest reaction layers in dynamic simulated body fluid at 72 h. The results indicated higher in vitro bioactivity for single-alkali glasses than mixed-alkali glasses in simulated body fluid.

Comparing in vitro reactions of granules and microspheres of the same size fraction and composition in dynamic TRIS buffer flow suggested that microspheres reacted more evenly than irregular granules. This difference was likely due to the microspheres’ more controlled size and packing degree. More developed calcium phosphate (CaP) layers were detected on the S53P4 spheres than on the granules. A shrinking core model approach was tested to calculate the release of the network modifier ions as a function of dissolution time for the microspheres of S53P4 and S53P4 with 5 mol% of K2O substituted for Na2O. The model provided an appropriate fit with the experimental data. These results are encouraging when testing approaches for predicting bioactive glass dissolution mechanisms. In the future, modelling will likely aid in developing new bioactive glass compositions and, thus, reduce the number of in vitro and in vivo studies.

Low strength limits the application of bioactive glass scaffolds to repair defects in load-bearing bones. Sintering of amorphous, porous S53P4 scaffolds with a strength suitable for handling in surgical applications was successfully conducted in a laboratory furnace. The sintering parameters were optimised by achieving a scaffold porosity of 50% and a compressive strength comparable to cancellous bone (5 MPa). Long-term dynamic dissolution experiments in TRIS buffer and simulated body fluid were carried out to study the in vitro properties of these sintered S53P4 scaffolds. These results indicated that the scaffolds gradually dissolved and formed reaction layers, while CaP strengthened the necks between the joined granules in simulated body fluid. These findings pave the way for the development of new S53P4 hot-worked products with enhanced strength and dissolution properties for future clinical applications.