Applications of scanning thermal microscopy in the analysis of the geometry of patterned structures

In this article, the results of the scanning thermal microscopy (SThM) analysis of artificial structures such as microchip surfaces and solar cell contacts are presented. It is shown that in the absence of surface roughness the SThM can be used to obtain reliable material contrast images. However, roughness and other topographical features can influence the thermal data in a strong way. It is illustrated that this effect can be partially removed by using the neural network approach for modelling the thermal signal using the topography data. The illustration of this approach is presented in the analysis of the geometry of the examples selected in this article. Copyright (C) 2006 John Wiley & Sons, Ltd.