No abstract

There has been a resurgence of interest in alpha-tricalcium phosphate (α-TCP), with use in cements, polymer composites and in bi- and tri-phasic calcium phosphate bone grafts. The simplest and most established method for preparing α-TCP is the solid state reaction of monetite (CaHPO4) and calcium carbonate at high temperatures, followed by quenching. In this study, the effect of the chemical composition of reagents used in the synthesis of α-TCP on the local structure of the final product is reported and findings previously reported pertaining to the phase composition and stability are also corroborated. Chemical impurities in the monetite reagents were identified and could be correlated to the calcium phosphate products formed; magnesium impurities favoured the formation of β-TCP, whereas single phase α-TCP was favoured when magnesium levels were low. Monetite synthesised in-house exhibited a high level of chemical purity; when this source was used to produce an α-TCP sample, the α-polymorph could be obtained by both quenching and by cooling to room temperature in the furnace at rates between 1 and 10°C/min, thereby simplifying the synthesis process. It was only when impurities were minimised that the 12 phosphorus environments in the α-TCP structure could be resolved by (31)P nuclear magnetic resonance; samples containing chemical impurity showed differing degrees of line-broadening. Reagent purity should therefore be considered a priority when synthesising/characterising the α-polymorph of TCP.

High-purity (SupT) and reagent-grade (ST), stoichiometric and silicate-containing α-tricalcium phosphate (α-TCP: ST0/SupT0 and Si-TCP x=0.10: ST10/SupT10) were prepared by solid-state reaction based on the substitution mechanism Ca3(PO4)(2-x)(SiO4)x. Samples were determined to be phase pure by X-ray diffraction (XRD), and Rietveld analysis performed on the XRD data confirmed inclusion of Si in the α-TCP structure as determined by increases in unit cell parameters; particularly marked increases in the b-axis and β-angle were observed. X-ray fluorescence (XRF) confirmed the presence of expected levels of Si in Si-TCP compositions as well as significant levels of impurities (Mg, Al and Fe) present in all ST samples; SupT samples showed both expected levels of Si and a high degree of purity. Phosphorus ((31)P) magic-angle-spinning solid-state nuclear magnetic resonance (MAS NMR) measurements revealed that the high-purity reagents used in the synthesis of SupT0 can resolve the 12 expected peaks in the (31)P spectrum of α-TCP compared to the low-purity ST0 that showed significant spectral line broadening; line broadening was also observed with the inclusion of Si which is indicative of induced structural disorder. Silicon ((29)Si) MAS NMR was also performed on both Si-TCP samples which revealed Q(0) species of Si with additional Si Q(1)/Q(2) species that may indicate a potential charge-balancing mechanism involving the inclusion of disilicate groups; additional Q(4) Si species were also observed, but only for ST10. Heating and cooling rates were briefly investigated by (31)P MAS NMR which showed no significant line broadening other than that associated with the emergence of β-TCP which was only realised with the reagent-grade sample ST0. This study provides an insight into the structural effects of Si-substitution in α-TCP and could provide a basis for understanding how substitution affects the physicochemical properties of the material.