High-resolution metal replication of macromolecules

Abstract

By adjustment of various parameters affecting crystallite size in thin metal replica films used for contrasting biological macromolecules for electron microscopy, improvements in the level of image information retrieved are demonstrated. Mass thickness, replica metal type, substrate temperature and composition are found to be important determinants of replica quality and conditions are discussed and tabulated for optimizing these. Typical correction curves for replicas are given for platinum and tungsten, from which approximate true dimensions may be calculated from measurements of unknown replicas. Deviation of evaporant sources from ideality are quantitated for both heated filament and electron gun sources. Metal replicas exposed to 80 kV electron beams long enough receive 10(6)e/nm2 show no observable changes other than those attributable to contamination as a function of time after the first few seconds. A variety of artifacts occasionally produced in replicas are discussed in relation to image interpretation.