Surface topography analysis with application of roughness area dependence method.

At the moment the root mean square roughness (R) is the most commonly used parameter for quantitative description of surface properties. However, this parameter has one main disadvantage: for its calculation only height variations of surface profile are used which are then represented by a single number. To eliminate this restriction authors of the paper have developed a surface analysis method which is based on roughness calculation in the function of gradually increasing dimensions of a sampling area. By setting proper measurement parameters and further data processing, from R dependence on sampling area plot size there is a possibility to obtain more useful, additional information about specific surface properties than using the single roughness value. Roughness area dependence plots, obtained from AFM images, were analyzed to study the influence of different growth parameters on surface properties of GaN layers and AlGaN/GaN heterostructures grown on sapphire and silicon substrates by Metal Organic Chemical Vapor Deposition (MOVPE) epitaxy. Although the method was used to characterize the semiconductor material in micrometer range, it can be applied also for any topography imaging technique in wide scale ranges.