Abstract
X-ray Photoelectron Spectra (XPS) are unusual because of the fact that the background changes rapidly in the vicinity of the peak position and with an amount which is a significant proportion of the peak area. Correct estimation of peak and background intensity is therefore necessary for quantitative analysis. Detailed theoretical interpretation of background shape may also open new opportunities for modelling of the nearsurface composition gradients. The problem is made difficult because neither the peak nor the background have a known shape and their separation requires the photoemission process to be treated as a sequence of well separated events. This thesis explores the background subtraction programs which are available and attempts to codify their limitations and to understand their points of overlap. The action of these programs is then examined using spectra of representative elements drawn from all parts of the periodic table. An effort is made to establish general rules and trends which will assist in the choice of appropriate programs for use by the analytical community. In this thesis the literature is first reviewed giving the history of background handling strategies in XPS and the basis used to establish computer programs by which they are achieved. Fits are then made, using different programs, on various elements which have very different types of background both in magnitude and in shape. The results concerning the completed data set are discussed and a general trend is extracted which can help in deriving the correct fitting choices for the interpretation of XPS data.