abstract
- PURPOSE: Transmission electron microscopy (TEM) is widely used to study the ultrastructure of bone. The mineral of bone occurs as polycrystalline mineral plates about 3 to 6 nm in thickness. A problem in using TEM to make quantitative analyses of bone is that the orientation of the plates with respect to the plane of the section being imaged is expected to affect their apparent thickness. The purpose of this study was to test if this was true, if the apparent thickness of plates changed substantially as a result of tilt of the section. METHODS: We prepared TEM sections of samples of cortical human bone by ion beam milling, orienting one section parallel to the collagen fibril axes and one perpendicular to them. We obtained TEM bright field and HAADF images of these sections, tilting the sections up to ± 20° at 2° intervals and measuring the apparent thickness of individual mineral platelets at each angle of tilt. RESULTS: Thickness appears to double as section is tilted ± 20°. True thickness of plates is determined by tilting the section along an axis parallel to the plate orientation and determining the minimum apparent thickness. However, as plates are tilted away from minimum-thickness orientation, they become less well-resolved, disappearing when tilted more than 20°. We therefore also measured apparent thickness of only the darkest (most electron scattering) plate images in an untilted section and obtained the same average thickness as that obtained by tilting. CONCLUSION: We conclude that tilting of the section is not necessary to obtain an accurate measurement of the thickness of mineral plates.