Effect of specimen processing for transmission electron microscopy on lattice spacing variation in Si specimens

•Lattice spacing in Si is well used for magnification calibration of TEM.•Lattice spacing in Si specimens for TEM may vary depending on specimen processing method.•The lattice spacing at the outermost region of the specimen edge tends to increase.•The variations can be component of the measurement uncertainty in TEM.•The Si lattice spacing is not always suitable for use as an SI-traceable reference. Variation in the (220) lattice spacing of Si due to specimen processing for transmission electron microscopy (TEM) was experimentally evaluated by comparing the measured lattice spacings of crystalline specimens processed by crushing, mechanical polishing only, and combined mechanical and Ar ion polishing. Although distinct variation in the (220) lattice spacing between the Si specimens processed by crushing and by mechanical polishing only is not observed, the (220) lattice spacing of specimens prepared by combined mechanical and Ar ion polishing imply increasing tendency with increasing Ar ion beam irradiation time. Moreover, the (220) lattice spacing measured from the outermost region of the specimen edge tends to be approximately 3% to 5% larger than that measured from the inner region, irrespective of the processing method. These results demonstrate that differences in the processing conditions of Si specimens and in the measurement location of the Si lattice spacing can be major component of the measurement uncertainty in sub-nanometer metrology using TEM with magnification calibration by the Si lattice spacing. When attempting to apply the lattice spacing of Si as a reference with traceability to the International System of Units for TEM magnification calibration in sub-nanometer metrology, the results suggest that the effect of specimen processing on variation in the lattice spacing is not negligible.