No abstract

No abstract

The main problems associated with matrix-assisted laser desorption/ionization (MALDI) quantitative analysis are poor shot-to-shot reproducibility, crystal inhomogeneity, signal degradation, and data acquisition system nonlinearity. These problems are addressed in this paper. The instrumental setup was modified to obtain required linearity and measurement accuracy over a dynamic range

Ne wcastle-upon-Tyne, Ne wcastle-upon-Tyne, NE 1 7RU, United KingdomVarlous data analysls technlques of Importance In X-ray photoelectron spectroscopy (XPS) are tjlscussed. Inelastlc background determlnatlon and subtractlon and the use of derlvatlve and dlfference spectra are illustrated wlth particular blas toward 0 1s and C 1s spectra obtalned from carbon flbers. Derlvatlve spe0tra, as a relatlvcsly qulck method of resolutlon enhancement, are shown to provlde a fairly accurate, though qualltatlvc? guide to the makeup of convolutod peak envelopes. The llmportance of spoctral allgnment and normallration in difference spectra are Illustrated, and procedures for determlnlng optimal resuks are shown for complex C 1s spectra. Correct {spectral allgnment is found to be more Important than correct normallration for dlfference spectra.A large amount of useful information can be obtained from the analysis of digital X,-ray photoelectron spectroscopic (XPS) data using either large! main frame computers or small microprocessor computem. The most effective method is probably that of nonlinear least-squares curve fitting (which is quite different from curve deconvolution). Curve fitting can be used for valence and core level studies, giving accurate information about the peak parameters, such as position, width, and intensity, with a gauge of their error and a statistical estimation of the goodness of the overall fit, togeth.er with the inclusion of the effects of radiation satellite peaks.Other particularly useful techniques include spectral smoothing, inelastic background subtraction, and the use of difference and derivative spectra. We hiave found the latter techniques to be extremely useful in the ;analysis of XPS data where the changes between spectra are small, such as in t:he spectra of carbon fibe:r surfaces ( 1 , Z ) .The best approach to data analysis requires the use of as many methods as possible, as each gives information which can confirm or cast d.oubt on conclusions from the otber methods.The aim of this palper is to discuss3 the application of nonlinear background ;subtraction, derivative, and difference spectra in XPS data amalysis.

No abstract