Analysis of X-ray Photoelectron Spectra through Their Even Derivatives

Abstract

The analysis of organic and inorganic surfaces can be carried out very effectively with the aid of x-ray photoelectron spectroscopy. In many cases, however, the presently available methods and techniques for data treatment resolutions are not suitable for the qualitative and quantitative identification of the various forms of a given atom on the same surface. The number of components and a good approximation of their original position in the composite curve must be known to use the available curve fitting procedures, otherwise the evaluation can be unreliable. It is suggested that the second and higher even derivatives of the composite could provide these data. The possibility of applying even derivatives of composite curves in combination with a nonlinear least square curve fitting program was investigated. It was found that depending on the noise background of the spectra, the resolution could be improved through this method. The resolution is dependent on the half-width of the component curves, their separation, and ratio. Both Gaussian and Lorentzian curves can be resolved, but the resolution of the latter is easier. The resolution is increasing with higher derivatives; however, increased smoothing must be applied at each step to neutralize the influence of the noise background.