On the selection of an integration limit for quantitative XPS analysis

Abstract Difficulties are frequently encountered in quantitative XPS analysis while selcting a precise background level for measurement of the area under a photoelectron peak. For this, linear background subtraction is accepted to be a satisfactory method where an integration limit is selected at a constant multiple of the inherent line width. It is pointed out, however, that an improper selection of linear background may lead to wrong quantification. In the present investigation, selection of a suitable background level has been achieved using the range-dependent property of variance of a given distribution. Systematic effects contributing to the line intensities are taken into account. Experimental measurements have been made of the relative intensities of the prominent photoelectron lines of gold, silver and copper. These are compared with those obtained from simple model calculations. An agreement within 6% is reached between the calculated and the measured values of intensities. A comparative study has also been made with the existing methods of background subtraction. Finally, the present technique is used to determine the stoichiometry of Al 2 O 3 .