Bioactive glasses as reactive biomaterials: Dissolution-driven structure–function relationships and predictive modelling
Acta Biomaterialia, vol.8, pp.1-15, 2026 (Scopus)
- Publication Type: Article / Review
- Volume: 8
- Publication Date: 2026
- Doi Number: 10.1016/j.actbio.2026.08.005
- Journal Name: Acta Biomaterialia
- Journal Indexes: Scopus
- Page Numbers: pp.1-15
- Yıldız Technical University Affiliated: Yes
Abstract
Bioactive glasses (BGs) are reactive biomaterials whose biological effects arise from dissolution: ion exchange, pH evolution, surface-layer formation and calcium phosphate precipitation. Their performance is therefore shaped not by nominal composition alone, but by glass structure, processing, crystallinity, surface area, form factor, medium and test protocol. This critical narrative review reframes BG bioactivity as a time-resolved structure–dissolution–function problem and examines which predictive claims are currently justified. We distinguish direct BG evidence from mechanistically relevant non-BG glass studies and modelling examples from adjacent fields. We discuss how network connectivity, Qⁿ speciation, phosphate and borate structure, modifier identity, structural heterogeneity and accessible surface area govern ion release, pH, surface-layer evolution and mineralisation. Silicate, borate/borosilicate, phosphate-based and sol–gel/mesoporous systems are compared as distinct dissolution programmes rather than fixed bioactivity labels. Atomistic simulations are considered useful sources of structural descriptors when checked against experimental dissolution and biological data. Across these systems, ion-release profiles, pH and supersaturation trajectories, mineralisation kinetics, cytocompatibility windows, antimicrobial or angiogenic cues and mechanical retention can only be interpreted when exposure protocols are reported. Current modelling evidence is strongest for physicochemical outputs, particularly dissolution and ion release under defined conditions. Predicting pH, mineralisation, and cell response requires measured exposure variables, biological metadata, and external validation; in vivo performance remains indirect without paired datasets. We conclude that progress depends less on more complex models than on better-reported experiments, clear uncertainty and cautious claim boundaries.