Penn State researchers have refined a computational model that simulates how silicon carbide transforms from solid to vapor during crystal growth, potentially helping manufacturers produce higher-quality material for advanced electronics.
The study, led by Yun Kyung Shin, research professor of mechanical engineering at Penn State, was published in The Journal of Physical Chemistry C. It represents the first publication from Penn State's Silicon Carbide Innovation Alliance, an industry-university partnership.
Silicon carbide is a crystalline material used in high-power electronics for electric vehicles and clean-energy systems. Unlike other crystals grown by melting and cooling, silicon carbide grows through physical vapor transport, where solid material heats until it vaporizes directly, then the vapor settles on a seed crystal to build a larger one.
Because this process occurs at extreme temperatures inside a closed furnace, researchers cannot directly observe the early stages of sublimation or identify which molecules are present in the vapor. "Direct in situ experimental observation inside a closed PVT furnace is nearly impossible, leaving a critical gap in our understanding of early-stage sublimation and vapor dynamics," Shin said.
The team refined ReaxFF, a computational method that uses mathematical rules to calculate how atoms interact and simulate chemical reactions. Adri van Duin, Evan Pugh University professor of mechanical engineering, co-invented and developed ReaxFF.
The researchers reparameterized the force field using quantum mechanical calculations to study sublimation pathways that differ from earlier work focused on oxidation. The refined model let researchers track silicon- and carbon-containing molecules as silicon carbide heats and vaporizes. Results closely matched detailed quantum calculations and laboratory experiments confirmed that solid material left behind becomes increasingly carbon-rich.
Simulations showed vapor composition changes with temperature. At lower temperatures, silicon-rich species escape more easily due to lower activation barriers. Vapor that is too silicon-rich can produce silicon droplets or large defects, while excess carbon can cause graphitization or unwanted crystal structures.
The team calculated temperature-dependent sublimation reaction rates that could incorporate into furnace-scale computational fluid dynamics models. Combining atom-scale reaction data with furnace-scale simulations could enable virtual growth experiments. "This allows manufacturers to optimize furnace geometries and thermal recipes through simulations, thereby lowering the reliance on costly and time-consuming trial-and-error growth processes," Shin said.
A complementary study by Morgan Advanced Materials researchers, published in the Journal of Crystal Growth, developed a thermodynamic model of equilibrium vapor composition during growth. Together the studies examined silicon carbide sublimation across scales, from individual atoms to the overall vapor environment.