Expansion of statistical overlap theory by maximum entropy method

ANYL 280

Yuping Williamson, williamsonyv@yahoo.com and Joe M Davis. Department of Chemistry and Biochemistry, Southern Illinois University at Carbondale, 1245 Lincoln Dr, 224C, Carbondale, IL 62901
The original formulation of statistical overlap theory (SOT) began with Davis and Gidding's work on point-process analysis. This theory has since opened a new window for a deeper understanding of chromatographic separations with regards to multi-component mixtures. A limitation that discourages this theory from being used broadly is the prerequisite of the frequency, or the statistical distribution of the number of single component peaks (SCP) per unit time. An assumption-free regimen for the frequency determination is proposed in this presentation. The maximum entropy method (MEM) is used to relate the number of peak maximum in different region of a separation to the optimal frequency. In the process, the number of mixture compounds is determined. MEM is a model-free regimen that is suitable to solve problems involving the presence of incomplete information (in this case, the SCPs obscured by overlap). The work can greatly broaden the practical use of SOT.

 

Analytical Approaches: Separation Science
1:30 PM-5:00 PM, Wednesday, 13 September 2006 Moscone Center -- Room 130, Oral

Sci-Mix
8:00 PM-10:00 PM, Monday, 11 September 2006 Moscone Center -- Hall D, Sci-Mix

Division of Analytical Chemistry

The 232nd ACS National Meeting, San Francisco, CA, September 10-14, 2006