6.4 Spray Pyrolysis Deposition

The exact deposition process for spray pyrolysis is not yet known. As explained in Chapter 5, there are suggestions that the process follows a CVD-like deposition as liquid droplets approach the surface and fully convert to their vapor form due to heating. However, others suggest that the process is more a liquid layer-by-layer thin film deposition. Two different topography simulations are performed. The first simulation uses ethanol-based droplets with a YSZ precursor which is atomized using an ESD system and appears to deposit on the silicon substrate as a liquid film. The second simulation uses a water solution with a SnO$ _2$ precursor, atomized with an air blast atomizer which appears to deposit on the silicon substrate in the form of a CVD process.

It is likely that the deposition of the liquid film using an ESD system [171] is due to relatively large particles being atomized and transported with two push forces, gravity and the electric force, towards the silicon surface. Many droplets still have most of their volume in tact as they reach the thermal zone, allowing them to come in contact with the heated silicon in order to deposit a thin disk-shaped film. The velocity of particles which reach the thermal zone in an ESD system is also shown to be higher than those using a PSD system, due to the additional electric force. Therefore, the introduction of the retardant thermophoredic force is not enough to significantly slow down and evaporate the large particles present.

However, the PSD process appears to deposit in the form of a CVD-like system [64]. There are two explanations why this setup does not appear to generate a liquid thin film deposition. The first is the lack of additional pushing force to deliver unevaporated droplets to the silicon surface, meaning they rely solely on the force of gravity and the initial velocity to bring them to the surface. As will be shown, the retardant Stokes force is much stronger than the force of gravity, making the droplet quickly reach a relatively slow terminal velocity. As the droplet reaches the thermal zone, an additional retardant force causes it to slow down even further, making it spend a long time in a hot environment and subsequently evaporate. An additional explanation for the CVD-like deposition is the atomizer setup used. The atomized droplets have a relatively low size, a mean radius of 2.5$ \mu m$, which is much smaller than the ESD system from [171]. Table 6.1 shows the relative parameters for the solutions and droplet used in the simulations for both the ESD system (ethanol and YSZ) and the PSD system (water and SnO$ _2$).

Table 6.1: Characteristics of the precursor solutions used for the simulations.
Droplet properties
  Ethanol Water
Droplet conductivity $ \left(\kappa_{d}\right)$ 0.19 $ W\slash m\cdot K$ 0.609 $ W\slash m\cdot K$
Droplet density $ \left(\rho_{d}\right)$ 789 $ kg\slash m^{3}$ 998 $ kg\slash m^{3}$
Absolute viscosity $ \left(\eta_d\right)$ 0.00116 $ N\cdot s\slash m^2$ 0.01 $ N\cdot s\slash m^2$
Surface tension $ \left(\gamma\right)$ 0.022 $ N\slash m$ 0.072 $ N\slash m$
Permittivity $ \left(\epsilon\right)$ 25 $ \epsilon_{0}$ 80.1 $ \epsilon_{0}$
Molar weight $ \left(M_W\right)$ 46.1 $ g\slash mole$ 18.0 $ g\slash mole$
Average diffusion coefficient ($ D_{v,f}$) 6.314$ \times $10 $ ^{-22}\times{T}\slash{r}$ 7.325$ \times $10 $ ^{-23}\times{T}\slash{r}$
Saturation vapor pressure (SVP) 5380 $ Pa$ 2340 $ Pa$
Boiling point (1atm) 351.5 $ K$ 373 $ K$
Maximum droplet charge $ \left(q_{max}\right)$ 1.11$ \times $10 $ ^{-5}\times r^{{3}\slash{2}}$ 2.01$ \times $10 $ ^{-5}\times r^{{3}\slash{2}}$
Air$ \slash$ambient properties
Air viscosity $ \left(\eta_a\right)$ 2.2$ \times $10 $ ^{-5} N\cdot s\slash m^2$
Air density $ \left(\rho_{a}\right)$ 1.29 $ kg\slash m^{3}$
Air thermal conductivity $ \left(\kappa_{a}\right)$ 0.025 $ W\slash m\cdot K$
Gas constant $ \left(R\right)$ 8.3144621 $ m\cdot N\slash K\cdot mol$
Simulation properties
Minimum droplet radius $ \left(r_{min}\right)$ 2.5$ \mu m$ 1.5$ \mu m$
Maximum droplet radius $ \left(r_{max}\right)$ 55$ \mu m$ 5$ \mu m$
Atomizer height $ \left(H\right)$ 270mm $ H_x$=200mm, $ H_y$=100mm
Droplet distribution angle $ \left(\theta_d\right)$ 45 $ ^{\textrm {o}}$ 12 $ ^{\textrm {o}}$
Flow rate 2.8 $ ml\slash h$ $ <$ 3.1 $ ml\slash h$
Electric potential $ \left(\Phi_0\right)$ 10$ kV$ -
$ K_V$ $ \sim 1 $ -
Evaporation parameter $ \left(q_0\right)$ $ \sim $373 $ \mu m^2\slash s\cdot K$ $ \sim $88 $ \mu m^2\slash s\cdot K$
Evaporation parameter $ \left(q_1\right)$ $ \sim $8.91$ \times $10 $ ^{-5} \slash \mu m$ $ \sim $4.3$ \times $10 $ ^{-3} \slash \mu m$
Temperature $ \left(T\right)$ 523$ K$ 673$ K$
Temperature gradient $ \left(\nabla T\right)$ 100,000$ K\slash m$ 30,000$ K\slash m$


L. Filipovic: Topography Simulation of Novel Processing Techniques