(301f) Detailed Kinetic Modeling of Complex Reactions: Reaction Network and Parameter Estimation Issues

The considerable interest in molecule-based kinetic models for complex chemistries is motivated by the need to predict product properties as a function of chemical process conditions. This is because the molecular composition is an optimal starting point for the prediction of mixture properties. Reactivity is an especially significant property that can be discerned given a molecule's (and its reaction environment's) structure. Other properties fall into performance and environmental classes. Thus, the potential advantages of molecule-based modeling are clear. Less readily apparent, how¬ever, is that the development and operation of molecular models comes with a large requirement for model construction and solution time as well as foundational reactivity information.

The procurement of this reactivity information is generally the result of well-defined laboratory experiments with simple or model feeds as reactants. Even in these instances, the goal of obtaining intrinsic information can be obstructed from issues derived from the complexity of the reaction network. This is because experiments with pure-component starting materials generally become complex multi-component mixtures even at low conversions, which makes resolution of the reaction network and quantitative rate law parameters a non-trivial exercise.

This talk examines Ken Bischoff's contributions to two of these areas. The DelPlot technique is an extension of the classic selectivity (s) vs. conversion (x) analysis to sort products according to their rank (R), with primary products having R = 1, secondary products having R = 2, and so on. At the heart of the DelPlot technique is a set of plots of s/x^r vs. x, and reaction products having a non-zero intercept on a given plot are determined to have rank R = r. The establishment of the reaction network allows, in turn, the determination of the parameters of the reaction rate law. The Himmelblau-Jones-Bischoff technique is a remarkable insight that provides a rapid estimate of parameter values that can be used as a starting point for further optimization by other methods.