Heat transfer and mixing in initiated chemical vapor deposition analyzed by <i>in situ</i> gas composition sensing

When interpreting or modeling initiated chemical vapor deposition (iCVD) data, assumptions about vapor phase heat and mass transport are almost always required. However, empirically verifying such assumptions is difficult and rarely attempted. To address this gap, we utilize in situ gas composition measurements and heat transfer experiments to validate theoretical and computational models of the iCVD system. Gas composition is obtained using a Pirani pressure gauge, which we demonstrate to be effective in simple gas mixtures. Dimensional analysis shows that the parameter space describing heat and mass transport can be simplified significantly, with the degree of mixing in the reactor determined by the Peclet number (Pe) and the heat transfer regime determined by the Knudsen number (Kn). Multiphysics simulations indicate that the well-mixed assumption holds below Pe = 0.3. Residence time distribution measurements for argon and isopropanol support this conclusion. Next, we study heat transfer by measuring thermal diffusion from the filament at a range of pressures in argon and isopropanol. These data show that heat transfer is pressure independent above Kn = 0.4 and validates a first-principles model of heat transfer over the full range of Kn. To account for thermally driven pressure changes in batch mode, we define an “effective temperature” for the reactor and validate a model of it against pressure versus filament temperature data. Our findings provide a systematic framework for iCVD practitioners to estimate mixing and temperature profiles and assess the impact of design choices on both batch and continuous operation.