J.C. Elliott

Rietveld refinement of the crystallographic structure of human dental enamel apatites

Abstract Rietveld refinements using 12 sets of X-ray diffraction powder data from milligram samples of human dental enamel provide detailed information about the structure and composition of enamel apatite. The principal difference in atomic parameters between enamel apatite and Holly Springs hydroxylapatite is in O2, which is reflected in a reduction in the P-O2 bond length of 0.085 Å and PO4 volume by 3.6%. Modeling the hexad axis scattering with a single OH⁻ ion gives a 0.089 Å shift of the ion further away from the mirror plane at z = 1⁄4. The known distributed electron density along the hexad axis in enamel has been confirmed by direct comparison with synthetic hydroxylapatite. Although the CO32- ion position could not be determined directly, evidence for partial replacement of PO43- by CO32- ions came from an 8% diminution of the P site occupancy compared with that in stoichiometric hydroxylapatite. The observed reduction in the P-O2 bond length and PO4 volume in enamel is also consistent with this substitution. The loss of negative charge caused by CO32- replacing PO43- ions and loss of OH⁻ ions is compensated by loss of Ca2+ ions from Ca2 sites. The calculated density from the X-ray results is 3.021 g/cm3, in agreement with deductions from previous chemical analyses.

X-ray microtomography: nondestructive three-dimensional imaging for in vitro endodontic studies.

Article shows the application of a laboratory x-ray microtomography system, a miniaturized form of conventional computerized axial tomography, to the study of root canal morphologic characteristics and changes in the course of root canal treatment in extracted teeth. After reconstruction of the three-dimensional images, the IDL software package (Research Systems, Inc., Colorado) was used to obtain cross-sectional slices of the tooth and three-dimensional views of rendered surfaces of constant mineral density. The root canal systems and changes in these were imaged at a resolution (cubic voxel side-length) of approximately 40 microns.

Preparation and characterisation of monoclinic hydroxyapatite and its precipitated carbonate apatite intermediate

Five 100 g batches of a carbonate apatite (the intermediate) were produced by heating an aqueous slurry of CaCO3 and CaHPO4 with an overall Ca/P mole ratio of 5/3 with vigorous stirring. Each intermediate produced by boiling off water was heated in vacuum at 1100 degrees C to remove carbonate, then steamed at 900 degrees C to ensure complete hydroxylation. Comparison of calculated and observed X-ray diffraction patterns showed final products containing 50-100 wt% monoclinic hydroxyapatite (remainder hexagonal). Rietveld refinements in P6(3)/m gave structures similar to several hydroxyapatite standards, including NIST SRM 2910, although there was no evidence from X-ray diffraction that the latter was in the monoclinic form. Refinements from standards and final products were slightly different from published single crystal data for Holly Springs hydroxyapatite. This is attributed to known impurities in mineral hydroxyapatite and indicates that parameters from the Rietveld refinements are closer to the true values for pure hydroxyapatite. Rietveld refinements for intermediates showed small, but significant differences from the final product, the largest being in O1x, O2x and O(H)z. All P-O bond lengths were shorter than in the final product, resulting in a 3.2% lower PO4 tetrahedron volume. The occupancies of P and Ca(2) were reduced. These differences are attributed to partial replacement of PO4(3) by CO3(2-) ions.

Affinity binding phenomena of DNA onto apatite crystals

The effect of DNA on the crystal growth of hydroxyapatite (HAp) and its morphology was examined. X-ray diffraction patterns of DNA-containing apatites showed typical apatitic features. However, crystal growth was greatly inhibited in the presence of DNA during synthesis; particularly, the crystallinity in the a-axis direction decreased dramatically at low concentrations of DNA. The a- and c-axis dimensions of each precipitate were almost the same, which implies that DNA molecules affect only the crystal surface. CHN analysis clearly showed the presence of these elements, which increased with an increase in DNA concentration in the solution. Scanning electron micrographs of the precipitates formed in the presence of DNA showed typical needle-like crystals, with a decreased crystal size, especially width. Infrared absorption spectroscopy of the DNA-containing apatites showed that the 1630-1700 cm(-1) absorption band due to C=C and C=N stretching increased with an increase in DNA concentration during precipitation. The ESCA spectrum of HAp(DNA 1.0) shows N 1s and C 1s peaks that are absent and weak, respectively in HAp. 31P NMR spectroscopy revealed a weak peak at the base of the 31P peak from the PO4(3-) ions in the HAp crystals. This weak 31P peak had a small positive shift from the position found in native DNA which may be due to the phosphate backbone of adsorbed DNA. The apparent solubility of the HAps increased with an increase in DNA concentration. These results suggest that there is an affinity binding between apatite crystal and DNA molecules.

X-ray microtomography scanner using time-delay integration for elimination of ring artefacts in the reconstructed image

Abstract Most X-ray microtomography scanners work on the same principle as third-generation medical CT scanners, that is, the same point in each projection is measured by the same detector element. This leads to ring artefacts in the reconstructed image if the X-ray sensitivities of the individual detector elements, after any analytical correction, are not all identical. We have developed an X-ray microtomography scanner which uses the time-delay integration method of imaging with a CCD detector to average the characteristics of all the detector elements in each linear projection together. This has the added advantage of allowing specimens which are larger than the detector and X-ray field to be scanned. The device also uses a novel mechanical stage to “average out” inhomogeneities in the X-ray field. The results show that ring artefacts in microtomographic images are eliminated using this technique.

Synchrotron X-Ray Microtomographic Investigation of Mineral Concentrations at Micrometre Scale in Sound and Carious Enamel

Synchrotron X-ray microtomography (XMT) was used to measure the linear attenuation coefficient (LAC) for 1.9-µm sidelength voxels within approximal brown spot lesions and sound human enamel. XMT demonstrated three-dimensional features, notably sheets with ∼30 µm periodicity having low LAC, identified as regions of demineralization corresponding to Retzius lines. Quantitative three-dimensional measurements of mineral concentration, derived from LAC with assumption of a single model composition, were consistent with previous measurements of sound and carious enamel from microradiographic projections. The uncertainty in measurements of mineral concentration and mineral fraction volume was investigated by modelling enamel with a range of composition and component densities. This analysis showed that, although mineral concentration can be determined from LAC with an error of <0.2 g cm–3, the variation in pore fraction volume within caries lesions cannot be reliably determined from X-ray attenuation measurements alone.

Longitudinal Study of the Three-Dimensional Development of Subsurface Enamel Lesions during in vitro Demineralisation

A longitudinal study was made of the 3D development of subsurface enamel lesions in whole human molars. X-ray microtomography (XMT) was used to measure the 3D distribution of linear attenuation coefficients in the tissue at 8–15 stages during cumulative times of 36–107 days demineralisation through ∼1-mm-wide windows. Although lesion morphology was consistent with preferential anisotropic dissolution parallel to enamel prisms at the advancing front, detail (at a scale of ∼100 µm) varied in relation to exposed surface sites separated by <1 mm. The distribution of mineral in the most superficial region varied across the exposed face of each lesion. Within lesions, localised foci of low mineral concentration (at a scale of ∼200 µm) retained their general form through successive stages of demineralisation before coalescing. The most advanced regions within a lesion seemed to correspond with surface regions with lowest mineral concentration. These findings indicate that local variations in fractional pore volume of partially demineralised enamel influence the subsequent spatial development of lesions.

A comparison of the mineral content of enamel and dentine in human premolars and enamel pearls measured by X-ray microtomography.

Mineral content gradients in two composite enamel pearls from permanent human upper molars were measured by X-ray microtomography (XMT) at a resolution of 15-30 microns. This non-destructive microscopic technique was used to make 15-microns thick XMT slices with 100-microns separation through one pearl and 250-microns separation through the other. Average mineral contents were calculated from the linear absorption coefficients determined from regions of the XMT slices assuming the inorganic component to be calcium hydroxyapatite. These values were compared with similar XMT studies of coronal enamel and dentine of upper permanent premolars. A mineral content gradient in the pearls, reducing from the enamel surface to the amelodentinal junction, was found; this was similar to that observed in the coronal enamel of the upper premolar. The mineral contents in the surface and deeper enamel regions of the pearl were similar to those observed in premolar enamel. In contrast, the mineral content for the dentine of the pearl was greatest at the amelodentinal junction, i.e. the gradient was in the opposite direction to that observed in premolar dentine. These results suggest that the process of mineralization of the pearl dentine differs from that in permanent control dentine. In addition, gradients in enamel and dentine mineral contents reducing from the tip of the pearl to the base of the pearl were found.

Artefacts in X-ray microtomography of materials

Abstract X-ray microtomography is becoming an increasingly popular tool in the study of microstructure and failure mechanisms in biological and engineering materials, producing three-dimensional (3D) maps of the X-ray linear attenuation coefficient. Limitations of the technique are due to, for example, limited X-ray flux, use of polychromatic radiation (in laboratory systems), finite resolution, discrete sampling and X-ray scatter. These give rise to artefacts in the reconstructed image. Knowledge of these artefacts helps to distinguish them from real features and is important for optimising experimental design so as to minimise their effect on the results. To aid identification, artefacts were simulated computationally using an analytical phantom projection generator. Streak, ring, motion and beam hardening artefacts are considered as well as the results of errors in the centre of rotation and missing information in cone beam geometry. Examples of tomographic images of real materials specimens with motion and beam hardening artefacts as well as centring errors are also shown.

Rates of Mineral Loss in Human Enamel during in vitro Demineralization Perpendicular and Parallel to the Natural Surface

Human enamel is a structurally anisotropic material. The aim of this study was to investigate whether this structural anisotropy is reflected in the demineralization behaviour of enamel. Kinetics of demineralization of in vitro caries lesions with the direction of acid attack perpendicular to the natural surface of dental enamel from human premolar teeth were compared with kinetics when demineralized parallel to this surface. Pairs of enamel samples from the same tooth were demineralized under identical conditions. Loss of mineral with time was very nearly linear for both directions (consistent with the rate–controlling step being reaction at the advancing front rather than transport processes), but the perpendicular rate was, on average, about 14% higher than the parallel rate. The rate of demineralization parallel to the surface increased from the natural surface to the enamel–dentine junction by 10–25%, depending on sample. The origin of fine structure and slight departures from linearity in the loss of mineral with time plots are discussed. Mineral masses per unit area were determined from absorption of a 15–μm diameter X–ray beam using photon (AgKα) counting methods.