* Corresponding author. Biomaterials 21 (2000) 749}754 Microstructural dependence of Young's and shear moduli of P O glass reinforced hydroxyapatite for biomedical applications Maria A. Lopes, Rui F. Silva, Fernando J. Monteiro, Jose H D. Santos* INEB*Instituto de Engenharia Biome & dica, Laborato & rio de Biomateriais, Rua do Campo Alegre 823, 4150 Porto, Portugal Universidade do Porto, Faculdade de Engenharia da Universidade do Porto (FEUP), Departamento de Engenharia Metalu & rgica e Materiais, Rua dos Bragas, 4099 Porto Codex, Portugal Department of Ceramics and Glass Engineering, UIMC, Aveiro University, 3810-193 Aveiro, Portugal Received 30 June 1999; accepted 21 October 1999 Abstract P O glass reinforced hydroxyapatite composite materials were prepared through a liquid-phase sintering process. Secondary phases, - and -tricalcium phosphates (-TCP and -TCP), were formed in the microstructure of the composites, due to the reaction between the liquid glassy phase and the hydroxyapatite matrix. The dynamic Young's modulus (E) and shear modulus (G) of these composites were determined using an impulse excitation method. By applying the Duckworth}Knudsen equation, the elastic property results were correlated with the relative proportion of -TCP and -TCP phases and with the porosity percentage present in the microstructure. Glass reinforced hydroxyapatite composites showed lower Young's and shear moduli than unmodi"ed hydroxyapa- tite, mainly due to the presence of -TCP phase. The Duckworth}Knudsen model demonstrated an exponential dependence of E and G modulus with porosity and mathematical equations were derived for composite materials with porosity correction factors (b) of 4.04 and 4.11, respectively, indicating that porosity largely decreased both E and G moduli. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Young's modulus; Shear modulus; Duckworth}Knudsen model; Hydroxyapatite glass composites 1. Introduction Calcium phosphate ceramics and glass ceramics have received much attention as bone implant materials be- cause they bond chemically to natural bone. The selec- tion of a material for many biomedical applications is a compromise between biocompatibility and mechanical performance, namely when the implant material is under high loads. It is important to determine the elastic prop- erties of a material to evaluate the mechanical interaction between implant and surrounding tissues, and therefore to develop new materials that match more closely to bone in sti!ness [1,2]. Several techniques have been used to measure elastic properties of ceramics, namely static bending, resonance, ultrasonic and impulse excitation methods [3,4]. The last technique consists of measuring the fundamental reso- nant frequency of test specimens by exciting them mech- anically through a single-elastic strike with an impulse tool. This dynamic method shows di!erences from other techniques as it does not require continuous excitation, it is non-destructive, the specimen is subjected to minute strains and there is no requirement for complex support systems that demand elaborate set-ups or alignments. It is also suitable for testing specimens with complex geo- metry. P O glass reinforced hydroxyapatite composites have been developed with enhanced [5,6] biaxial bend- ing strength and fracture toughness, when compared with unmodi"ed sintered hydroxyapatite, HA, Ca (PO ) (OH) . Cell culture studies using a human osteosarcoma cell line have demonstrated that these composites are highly biocompatible, and their e!ect on the cell cycle and on the expression of multiple antigens have been reported previously [7,8]. In order to measure dynamic Young's and shear modulus of glass-reinforced HA composites, an impulse excitation of vibration method was used. The depen- dence of these two elastic properties on microstructural characteristics of the composites, such as relative propor- tion of secondary phases (-TCP and -TCP) and 0142-9612/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 9 9 ) 0 0 2 4 8 - 3