Date of Award
Dissertation - Restricted
Doctor of Philosophy (PhD)
Electrical and Computer Engineering
Thomas C. Ehlert
Michael D. Ryan
Thomas A. Bronikowski
Daniel T. Haworth
Thomas K. Ishii
Deconvolution is a technique used in many fields. In spectroscopy, deconvolution is used to unfold the broaden effect caused by the measuring instruments. Consequently, the resolution of the spectra will enhance. In electron impact (EI) ionization, deconvolution is used to totally or partially remove the Boltzmann type thermal electron energy spreading of the EI data.
In this study five deconvolution methods were investigated. They were serial division method, inverse convolver method, electron energy distribution difference method, iterative deconvolution method, and Fourier transform deconvolution method. An EAI QUAD 250 mass spectrometer was used to collect the EI data of calibrant gases and the high temperature copper fluorides. A PDP8/e minicomputer was interfaced to the mass spectrometer for data acquisition and signal averaging. The signal averaging data was then deconvoluted using a Sigma-9 computer.
Prior to the application of deconvolution to an experimental data, a synthetic data was used to study the iterative deconvolution method and the Fourier transform method. The effect of several parameters on the deconvolution result was identified.
In the study of experimental data, Fourier transform deconvolution and iterative method of data smoothing were used. The electron energy spreading, called , the instrument function (IF), was obtained from the second derivative of The IF obtained
from this mass spectrometer was different from the Boltzmann type of energy distribution. The unfolded spectra were obtained by deconvoluting the signal averaging EI data with the instrument function. From the unfolded spectra, the high temperature ionization potential and appearance potential of copper monofluoride were determined.