(653g) Graph Theory Applied to Plasma Chemical Reaction Engineering | AIChE

(653g) Graph Theory Applied to Plasma Chemical Reaction Engineering


Holmes, T. D. - Presenter, The University of Sheffield
Rothman, R., University of Sheffield
Zimmerman, W. B., University of Sheffield
This work has investigated the application of graph theory to plasma chemical reaction engineering in the following ways: Assembling a weighted directional graph from a set of plasma chemical kinetic reaction data for air, with the crucial step of setting the reactions as nodes. Network visualisation of the graph for probing the plasma chemical kinetic reaction system for useful reactions which could be enhanced (with catalysts for example), or troublesome reactions steps which could be supressed. Running Dijkstra’s shortest path algorithm on the graph between each two species nodes. Assessing how the graph could be used to provide useful engineering data, particularly which conditions, reactions, or species could be most useful in producing a desired outcome, such as a specific chemical.

It was found that the use of reaction-nodes combined with the use of some characteristic parameters allowed a large amount of key information on the air plasma chemical reaction system to be assessed simultaneously using a leading open source graph visualisation software (Gephi). From a connectivity matrix of runs of Dijkstra’s shortest path algorithm between each two species it could be seen which species had a higher potential to be formed than to be destroyed, and vice versa.

Further investigation into how the graph could be used to provide key information for reaction engineering lead to the development of a relatively simple algorithm: Optimal Condition Approaching via Reaction In Network Analysis (OCARINA). The algorithm analyses the graph system to suggest conditions could be most beneficial for the formation of a selected species. The predictions given by running OCARINA on a pre-existing air plasma reaction dataset display significant similarities to a well-known electric field strength regime for optimal ozone production. This could potentially be a useful aid in the selection of simulations and experiments to direct valuable time and resources.