The use of biomolecules in nature to direct the path of crystal growth leads to a degree of polymorph and morphology control that far surpasses anything currently accessible in a laboratory. In particular, single-crystal assemblies, with morphologies supremely tailored to their functional requirements, are found in a wide range of biological settings. It is not surprising that many researchers seek to mimic biomineralisation processes in order to fabricate new materials, but the scope for designing such materials would be greatly enhanced if we had a more robust and fundamental understanding of biomineralisation. In particular, a crucial feature is the extent to which nucleation of crystalline material is controlled: sometimes suppressed to avoid polycrystalline formations; other times templated to produce specific polymorphs or even facets; but always controlled.
Over the last 5–10 years, a great deal of information has emerged about the onset of crystal formation in biominerals, and this has forced a seed-change in the understanding of biomineral nucleation. It is now clear that the early stages of crystal formation follow a strongly non-classical nucleation pathway, with a number of distinct stages prior to the appearance of crystalline phases. The following discussion illustrates these stages with reference to CaCO3, since this is the best studied system, but we note that similar mechanisms are implicated in phosphates and other carbonates. It is the understanding of these stages, both through experiment and theory, that is the focus of this work programme, and that will subsequently be used to enable more comprehensive studies of nucleation and growth within the other themes as the grant matures.
Very recent work has shown that the first stage in calcium carbonate deposition is actually the formation of amorphous CaCO3 clusters from solution. This has a substantial affect on the nature of the solid deposition “reaction”, as shown in the adjacent figure. The clusters are present in both under- and super-saturated solutions, and appear to be a stable thermodynamic feature of CaCO3 solutions. The pre-critical clusters are about 2 nm in diameter, and contain ~75 formula units of CaCO3; however little is known about their intrinsic structure or hydration state, and there is currently no theory for their formation. Speciation (CO32–, HCO3– etc.) is important within these clusters, and the eventual crystal polymorph that forms is correlated with the pH of the solution; but how speciation manifests within the clusters is again unknown. These pre-critical clusters subsequently aggregate, and crystal nucleation finally occurs within the aggregate of amorphous nanoparticles. In biomineralisation, the aggregation is often within an insoluble organic matrix that acts as a scaffold for the microscale structure of the biomineral. Crystal nucleation is then carefully controlled within this environment, to ensure few nucleation events are able to sustain crystal growth fronts, and hence ensure highly oriented crystalline assemblies that behave as single crystals, yet still exhibit remarkable flexibility in porosity, form and surface curvature. We note that a closely related phenomenon occurs with polymer-induced liquid precursors (PILP), where synthetic polymers are used to stabilise amorphous, but dynamic, nanoclusters of CaCO3, which can subsequently be used to facilitate the growth of various biomineral-like materials.