(458d) Reducing Reaction Mechanism for Synthesis of TiO2 Nanoparticles

Authors: 
Mehta, M., Iowa State University
Raman, V., University of Texas at Austin
Fox, R. O., Iowa State University


Using computational fluid dynamics to model nanoparticle synthesis in chemical reactors has been proven to be an important tool for understanding and implementing coupling between transport processes and chemical reactions[1]. Accurate modeling of synthesis in chemical reactors involves a detailed reaction mechanism. In developing a chemical mechanism we propose a list of species and take into account all feasible reactions. This leads to a large and highly detailed reaction mechanism involving species and reactions that might not have a significant effect on the reaction path of the species of our interest. For example West et al.[2] have proposed a reaction mechanism for TiO2 nanoparticle production from TiCl4 that involves 25 species and 51 reactions. Computational fluid dynamics techniques are used to couple the reaction mechanism with the transport processes but due to the large number of species and reactions involved, coupling this chemistry with detailed flow solvers is computationally very expensive. Thus, for practical simulations reducing these highly detailed chemical kinetics to moderate size is very important.

Reducing the number of species and consequently the reactions is non-trivial. We have to make sure that taking out species from the mechanism does not lead to inaccurate estimation of the desired species in comparison to the detailed set. Finding out how sensitive our species of interest are with respect to the reaction rate constants will tell us which reactions are important for their formation. Thus, the reaction set can be reduced by taking into account only reactions that are highly sensitive to the formation of the desired species.

The objective of this work is to develop a sensitivity analysis tool that given a reaction mechanism set, complete with reaction rate constants, can give us the sensitivity of the species with respect to specified reactions. Thus, reducing the number of species that the flow solvers has to track to make simulating nanoparticle synthesis more feasible and computationally tractable. We carry out sensitivity analysis on the detailed reaction mechanism given by West et al.[3] for the combustion synthesis of TiO2 from TiCl4 that includes 30 species and 66 reactions.

References:

1. Rodney O. Fox, ?CFD models for analysis and design of chemical reactors,? Advances in Chemical Engineering, vol. 31, Ed. G. B. Marin, Elsevier, pp. 231-305 (2006).

2. R. H. West, M. S. Celnik, O. R. Inderwildi, M. Kraft, G. J. O. Beran, and W. H. Green, ?Toward a Comprehensive Model of the Synthesis of TiO2 Particles from TiCl4,? Ind. Eng. Chem. Res., 46(19):6147?6156 (2007).

3. R. H. West, R. A. Shirley, M. Kraft, C. F. Goldsmith, W. H. Green, ?A Detailed Kinetic model for Combustion Synthesis of Titania from TiCl4 ,? Preprint. http://como.cheng.cam.ac.uk/index.php?Page=Preprints&No=55