ID 234

Polydispersity of nanoparticles produced via flame spray pyrolysis

 

Abstract:

Population balance models are commonly utilized for the description of the behavior of particles in many indus- trial systems. The application of such method on the numerical representation of the Flame Spray Pyrolysis (FSP) process can be a good ally for the prediction of its final product characteristics. The complexities associated to the FSP, especially regarding the formation and the growth (such as agglomeration and sintering) of the particulate ma- terial, require the use of a bivariate approach to better describe the particle population. The solution of the resulting bivariate population balance equation (PBE), nevertheless, can be very costly depending on the chosen method, and, for that, the Direct Quadrature Method of Moments (DQMoM) is chosen. The type of solution generated is also a subject for analysis, as monodisperse cases, most commonly found in literature, describe a size distribution using only its mean value, what may mask the function homogeneity, among other characteristics. In that sense, the present work aims to study the generation of a polydisperse solution for this problem, by means of the use of Computational Fluid Dynamics (CFD) coupled with a population balance model to represent the nanoparticle formation and evolution inside a FSP reactor. An analysis of turbulent Schmidt number, used in the definition of the effective diffusion of the nanoparticles, shows an optimum value of 4.65 to be employed in the studied cases. Furthermore, experimental results obtained via the Brunauer-Emmett-Teller (BET) method are used for validation of the results with error of less than 10% for the first moment of the distribution at the highest portion of the domain.