Several methods of measuring the thickness of foils used in transmission electron microscopy are described. The first method utilises available crystallographic features, such as slip plane traces, which are present in some materials. It is shown that by tilting the foil to align the defect plane parallel with the electron beam, very accurate values may be obtained for the foil thickness measured in the beam direction. The technique is not subject to errors which may be incurred when the foil surface orientation is not precisely known. The second technique involves depositing contamination spots on top and bottom surfaces of the foil by electron irradiation. Tilting the foil through a known angle causes separation of the spots in the image plane from which the foil thickness may be calculated. The technique is shown to overestimate true metal thickness due to the presence of surface films and a contaminant aureole surrounding the spots which makes precise location of the spot-substrate interface difficult. The convergent beam diffraction technique has been investigated and found to be accurate providing certain experimental conditions can be met. The results demonstrate that multiple beam diffraction and absorption effects can limit the accuracy of the method and that such effects are most pronounced for low order planes. The above techniques are compared and it is shown that the crystallographic and convergent beam technique measure the thickness of metal in the beam direction. A final method utilises x-ray measurements where the x-ray intensity is a function of beam current and foil thickness. The former dependency makes it difficult to correlate x-ray intensities with a particular value of foil thickness. Characteristic and continuum ratios may vary with thickness due to absorption and fluorescence effects and are independent of beam current. The observations indicate that a linear relationship between x-ray path length (and hence absorption) and foil thickness does not always occur. Fluorescence effects are found to be negligible in the 18.8 stainless steel foils studied. One thickness technique was used to estimate precipitate populations in aluminium alloys aged to peak hardness. It is shown that the habit plane is a function of the foil's silver content and that the apparent precipitate population is a function of foil thickness. A model is given to explain this phenomenon.
|Date of Award||1982|