Why are natural minerals such as seashells so hard?

Materials such as seashells, teeth and bones often have remarkable physical properties that are far better than those of equivalent artificially produced substances. These biominerals contain brittle minerals and flexible polymers, and often show highly organised structures. Together, these features combine to give the correct level of strength and durability required for their applications.

To find out why these naturally occurring biominerals have superior features, a structural study was carried out at Diamond Light Source. A simple model system to replicate the composition of these natural materials was constructed using calcite crystals containing the amino acids, glycine and aspartic acid. The resulting crystal lattice was analysed using the High Resolution Powder Diffraction beamline (I11).

The amounts of amino acids incorporated within calcite (CaCO3) crystals was altered in the model system to explore the effect of these additives on the crystal lattice and the hardness of the resulting crystals. The team found that the amino acids were occluded within the lattice as individual molecules, and that the hardness rose as the amino acid content increased. Moreover, they found that the hardening effect came from the way that defects within the crystal – which become mobile when the crystal is indented - cleave the amino acids in their paths. This fascinating discovery could help to improve strategies to make synthetic durable versions of these types of materials.

Biominerals such as bones, teeth, and seashells, exhibit many features such as complex morphologies, hierarchical structures and properties optimised for their roles that make them a unique inspiration for materials design1. Of particular interest is the ability of organisms to generate materials with strength and toughness appropriate for structural applications from weak, brittle minerals such as calcium carbonate and calcium phosphate. This enhancement in mechanical properties can be attributed to the hierarchical structuring of many biominerals, and intriguingly, to their inorganic/organic composite structures. Even single crystal biominerals exhibit superior hardness due to the presence of biomacromolecules within the crystal lattice2.

EE001
Figure 1: The amount of amino acid occluded within the calcite crystals, [AA]inc, as a function of the initial concentration of amino acid in solution, [AA]sol, for Asp and Gly. Inset shows a calcite crystal with modified morphology precipitated in the presence of 20 mM Asp and [Ca2+] = 10mM.

This study employs a simple model system – calcite (CaCO3) single crystals containing the amino acids aspartic acid (Asp) or glycine (Gly) – to investigate the origin of these strengthening and toughening mechanisms. Precipitation of calcite in the presence of Asp or Gly led to changes in crystal morphologies (Fig 1, inset) and extremely efficient occlusion of these amino acids within the calcite crystal lattice, where Asp occlusion reached a maximum of 3.9 mol% and Gly 6.9 mol% (Fig. 1). Importantly, the occlusion level could be tuned according to the concentration of the amino acids present in the crystallisation solution, and solid state nuclear magnetic resonance (NMR) confirmed that the amino acids were incorporated as individual molecules rather than aggregates.


Figure 2: (a) Lattice distortions arising from the incorporation of Asp and Gly in calcite along the c-axis. (b) XRD peak broadening (FWHM) due to inhomogeneous strains induced by Asp incorporation. These graphs show that the lattice distortions continue to increase with the amount of amino acid incorporated, while the inhomogeneous strains approach constant values. This behaviour is consistent with a model in which the local strain fields (e) associated with the individual amino acid molecules (c) start to overlap as the spacing between the individual molecules become smaller (d).

The High Resolution Powder Diffraction beamline (I11) was then used to investigate the influence of the occluded amino acids on the crystal lattice using synchrotron X-ray powder diffraction (SXPD)3. Occlusion of both Asp and Gly resulted in anisotropic lattice expansion (Fig. 2a), where the lattice distortionswere about an order of magnitude greater along the c-axis than the a-axis, an effect that is consistent with the elastic anisotropy of calcite. Analysis of the peak broadening, FWHM (full-width at half-maximum) provided information on the inhomogeneous lattice strains arising from the occlusion of amino acid molecules, and showed that for both Asp and Gly, the broadening increased with increasing levels of occlusion to levels of 1.5 mol% Asp and 2 mol% Gly, before levelling-off or even decreasing again (Fig. 2b). This is consistent with a model in which individual molecules generate strain in the surrounding lattice. As the number of occluded molecules increases, the strain fields around the molecules begin to overlap such that the strain inhomogeneity decreases as the lattice distortions continue to increase (Fig. 2c).

Molecular dynamics simulations provided valuable insight into the atomistic interactions between the mineral and amino acids and showed that Asp2- molecules are occluded such that the carboxylic acid groups on Asp replace CO32- groups on adjacent carbonate planes with a very good fit. A less favourable fit is obtained with Gly0, which may account for the reduced efficiency of incorporation of Gly than Asp. Simulations were also performed to model the lattice distortions arising from occlusion of amino acid molecules, where these data provide a bridge between our model of amino acid incorporation in the lattice and the experimental SXPD data, and an excellent agreement with the experimental values was obtained.

EE003
Figure 3: Hardness versus occluded amino acid ([AA]inc) for calcite occluding Asp and Gly. The inset shows a scanning force image of plastic indentation in calcite.

Finally, the mechanical properties of the amino acid-containing calcite crystals were measured using nanoindentation under load conditions where cracking was suppressed and indentations were formed by plastic deformation4. The hardness of the calcite crystals increased significantly with the quantity of amino acid occluded (Fig. 3). Starting at 2.5 GPa, which is equivalent to values recorded for Iceland spar, a pure geological calcite, the hardness increased to 4.1 GPa for occlusion of both 2.2 mol% Asp and 6.9 mol% Gly, where this is comparable to biogenic calcite. The indentation modulus, in contrast, was insensitive to the amount of occluded amino acid, as expected given the small volume fraction of the occluded phase.

 

The fact that we can precisely tune the amounts of Asp and Gly within calcite single crystals across a wide range of compositions then enables us to systematically study the effect of additive occlusion on both the lattice distortions and the hardnesses of the composite crystals. This in turn allows us to explore the origin of the hardening effect. Since the nanoindentation measurements were performed in a regime where the calcite deforms plastically without cracking, the increase in hardness caused by the occluded amino acids can be attributed to increased resistance to motion of dislocations on the active slip planes. The amino acid molecules can impede dislocations when either the stress fields around the molecules interact with the stress fields of the dislocations, and/or when dislocations are blocked by the molecules themselves. However, while the peak width (inhomogeneous strains) level off or decrease at occlusion values of ≈ 1.5 mol%, the hardness continues to increase with occlusion levels. The hardness is also comparable for calcite crystals containing the same levels of Asp and Gly, despite the fact that they exhibit quite different distortions (see original article for data for Gly). These observations suggest that direct blocking of dislocation motion by the amino acid molecules is the dominant hardening mechanism.

This hypothesis was further tested by estimating the ‘cutting force’ required for a dislocation to shear the amino acids within the calcite lattice. When a dislocation moves through a crystal containing an array of amino acid molecules, it is pinned by the molecules, causing it to bow out. Since lattice diffusion is not observed in calcite at room temperature, the dislocation can only continue to move if it cuts the obstructing molecule. The cutting force required to shear the amino acids was estimated based on the experimental hardness data and the corresponding amino acid contents of the crystals. Values of around 1 nN were determined, which lies just below reported strengths of single covalent bonds (1.5 to 4 nN) and well above a typical ionic bond strength of 0.1 nN. Since the hardness represents an average of all of the events that inhibit dislocation motion, this result strongly suggests that the hardening effect of these occluded molecules comes from the force required to shear them.

This work provides new insights into the mechanical properties of inorganic/ organic nanocomposites and finally determines the origin of the hardening effects of small organic molecules within single crystals. These results are of particular significance to the mechanical properties of single crystal biominerals, and open up the possibility of using this strategy to tailor the mechanical properties of a wide range of materials.

References:

  1. Wegst, U. et al. Bioinspired structural materials. Nature Materials 14, 23-36 doi:10.1038/nmat4089 (2014).
  2. Kim, Y. et al. An artificial biomineral formed by incorporation of copolymer micelles in calcite crystals. Nature Materials 10, 890-896 doi: 10.1038/nmat3103 (2011).
  3. Weber, E. et al. Intracrystalline inclusions within single crystalline hosts: from biomineralization to bio-inspired crystal growth. CrystEngComm 17, 5873-5883 doi: 10.1039/c5ce00389j (2015).
  4. Oliver, W. et al. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of Materials Research 7, 1564-1583 doi: 10.1557/Jmr.1992.1564 (1992).

Funding acknowledgement:

This work was supported by an Engineering and Physical Sciences Research Council (EPSRC) Leadership Fellowship (EP/H005374/1), by an EPSRC Materials World Network grant (EP/J018589/1) and an EPSRC Programme Grant (EP/I001514/1) which funds the Materials Interface with Biology (MIB) consortium. We acknowledge Diamond Light Source for time on beamline I11 under commissioning time and proposal EE10137.

Corresponding authors:

Dr Yi-Yeoun Kim, School of Chemistry, University of Leeds, Y.Y.Kim@leeds.ac.uk; Professor Fiona C. Meldrum, School of Chemistry, University of Leeds, F.Meldrum@leeds.ac.uk

Related publication:

Kim Y, Carloni J, Demarchi B, Sparks D, Reid D, Kunitake M, Tang C, Duer M, Freeman C, Pokroy B, Penkman K, Harding J, Estroff L, Baker S, Meldrum F. Tuning hardness in calcite by incorporation of amino acids. Nature Materials 15, 903-910, doi: 10.1038/nmat4631 (2016).

Publication keywords:

Calcium carbonate; Biomineral; Mechanical properties; Bio-inspired