Direct Methods with Electron Microscope Information

Abstract

The purpose of this chapter is to examine the process of structure solution when the data have been partitioned into two sets: The purpose of this chapter is to examine the process of structure solution when the data have been partitioned into two sets:

• (1) The basis set {H} comprising phased reflections where the phase information comes from the Fourier transform of electron microscope images after suitable filtering. Usually the phases so derived correspond to intensities that have a significantly lower resolution than the diffraction data, and one must also remember that there are some significant sources of error in image data:

• At the highest resolutions, more sampled pixels are needed to resolve a detail between two points. Even though averaging over the repeat of the two-dimensional space lattice can be used to minimise the actual radiation dose to the specimen when recording the image, the damage induced by inelastic interactions between electron and sample can be problematic [1].

• Many two-dimensional crystals, especially proteins, contain a curvilinear paracrystalline distortion, probably because of the lipid matrix in which the proteins are embedded. Thus, although the observed electron diffraction pattern might extend to e. g. 3A, the Fourier transform of a micrograph from a similar area may vanish somewhere in the range from 10 to 6A. Lattice unbending has been used to restore the higher resolution information [2], but the actual amount of artefact involved is not known.

• There are slight variations of specimen height from the perspective of the microscope objective lens. Although a nearby correction can be made locally for the lens focus before the low-dose image is recorded, the actual transfer function of the micrograph is often unknown.