This thesis is devoted to the development of near-field scanning optical microscopy (NSOM) for aqueous solutions and to fractal characterization of steel surfaces with atomic force microscopy (AFM) and other methods. NSOM combines optical properties from a light microscope and the technique of scanning probe microscopy, SPM (invented in the early 1980’s). With an appropriately configured scanning quartz pipette coated with aluminum, an NSOM can be constructed to operate in aqueous solution for applications in biology. Many of the technical limitations associated with a scanning pipette were circumvented, by the help of a small modulation of the distance between the pipette and the sample. This alternating current (AC) method allows the pipette to be positioned very close to the sample surface and is robust in obtaining reproducible NSOM images in solution. This approach is also compatible with fluorescence imaging and fluorescence resonance energy transfer (FRET), and should further facilitate the use of NSOM in various areas of cell biology, where high resolution is considered to be critical. Technical details of this design, and a further characterization of the system, are discussed in the context of biological applications as well as fundamental limitations in comparison to other systems. Based on the current technology, it is concluded that better than 50-nm resolution should be achievable with this technique for fluorescence and FRET imaging of biological specimens. The second part of the thesis is devoted to fractal analysis of data from AFM and other microscopic methods, such as scanning electron microscopy (SEM), light microscopy (LM) and profilometer. Fractal analysis is often necessary for studying surfaces with scale-invariant roughness, as is the case for many pre-treated steel surfaces. However, fractal parameters are influenced by the finite-sized tip geometry of the AFM stylus, and the dependence on AFM tip radius and surface height magnitude is analyzed according to different fractal methods. The result shows that fractal dimension is in general underestimated when the tip size is in comparison with features on the surface which may explain why higher fractal dimensions are seldom reported in the literature. When applied to mechanically pre-treated stainless steel samples, fractal dimension can be correlated to tensile strength for single overlap joints. This was shown within a length scale of ~0.5 - 100 µm, for profilometer profiles, using the Fourier and the Hurst algorithms, and for light microscope images, using a texture algorithm. In addition, the magnitude of the surface roughness, a parameter not often considered in fractal analysis, was shown to correlate to the arithmetic average difference, Ra. Hence, traditional parameters such as Ra tell very little about the spatial distribution of elevation data, in contrast to fractal dimension, and are not as easily correlated to tensile strength. Moreover, fractal dimension is closely related to the surface contact angle for a fluid, a parameter also important for adhesion, both for instant strength and durability properties. By comparing contact angle and adhesive data with fractal dimension, calculated from AFM and profilometer data for the steel surfaces, it can be shown that the surfaces can be qualitatively ranked according to anticipated adhesive properties.
Luleå: Luleå tekniska universitet, 2000. , 59 p.