A new implementation of a multidimensional solver for studying nanoparticle synthesis in laminar flames is presented. The governing equations are convective-diffusive-reactive partial differential equations that are discretised using the finite volume method. Detailed chemical source terms and transport coefficients are used to close the equations. The implementation of these governing equations is discussed and the numerical algorithm used to solve them is presented. The new solver is verified against analytic solutions and numerical solutions from 1D models for counterflow diffusion flames.

The new solver was used to calculate the flame location, shape and temperature of laminar premixed ethylene jet-wall stagnation flames when the equivalence ratio, exit gas velocity and burner-plate separation distance are varied. The simulation results were compared to new experimental 2D measurements of CH* chemiluminescence and temperature. The 2D simulations showed excellent agreement, and correctly predicted the flame shape, location and temperature as the experimental conditions were varied.

The new solver was used to study growth of inorganic nanoparticles in premixed, jet-wall stagnation flames. Titanium dioxide, also known as titania and TiO2, is a white powder than has many uses as a pigment, including in paper and cosmetics, and was selected as the system to apply the new solver. TiO2 nanoparticles formed from titanium tetraisopropoxide (TTIP) were simulated using a two step methodology, which enabled insight into the variations of particle properties as a function of the deposition radius. Two different TTIP loadings (280 and 560~ppm) were studied in two flames, a lean flame (equivalence ratio 0.35) and a stoichiometric flame (equivalence ratio 1.0). First, the growth of particles was described with a spherical particle model fully coupled to the conservation equations of chemically reacting flow. Second, particle trajectories were extracted from the 2D simulations and post-processed using a detailed particle model solved with a stochastic numerical method. The simulation produced gas phase predictions of flame location that are in good agreement with available literature. The particle morphologies and size distributions were examined and found to be dependent on the deposition radius. Particles began to have different size distributions at a deposition radius of approximately one and a half times the nozzle radius (1.0 cm), which should be kept in mind when synthesising and modelling nanoparticles for novel applications. This coincided with the growth of total residence time along particle trajectories. It is suggested that experiments critically examine the radially uniformity of deposited particles do not affect the performance for their intended application.