This technique really needs nanocrystals, usually 1–2µm in diameter, but less than 400nm thick. These can be grown by similar crystallisation techniques to those for X-ray crystallography and even if the conditions will not produce samples suitable for X-ray crystallography, they can actually contain useful nanocrystals for Micro-ED. The first issue is finding these as they aren’t easily resolved in a light microscope. However, screening of samples using negative staining EM is useful though the stain can cause deterioration of the crystals. It is also possible to produce nanocrystals by breaking down larger crystals, by sonication, vortexing, or even crushing.

Vitrified small crystal of a fluorobenzene complex sitting on the carbon film of a grid.

The cryo-FIB has been used to thin larger crystals to make lamellae. These samples produce exceptional quality diffraction patterns and structures of < 1Å have been obtained using them. The milling also removes the thick ice that can sit around the crystals and dampen high-resolution spots in the patterns.

Lipidic cubic phase or lipid bicelles have been used for proteins that need a lipid environment in which to crystallise. Proteins can diffuse and form small crystals within the continuous lipid bilayer. Membrane proteins are well suited to such techniques. These can be very viscous and specific dilution techniques or cryo-FIB have been used to make samples suitable for Micro-ED.