Optimized Nanosecond Laser Cutting of SiC Wafers: A Simulation‐Driven Approach

Abstract This study investigates the interaction of a 1064 nm nanosecond laser on 350 µm thick Silicon Carbide (SiC) wafers using optimized parameters derived from a simulation model and artificial neural networks (ANN). A decomposition temperature of 2800 K is considered for simulating effective laser ablation. Processing maps are generated for ablation width, depth, aspect ratio, and efficiency, using laser power and scanning speed as key inputs. The optimized process resulted in an ablation width of ≈28 µm, depth of ≈350 µm, aspect ratio of ≈12, and an ablation efficiency of ≈1700 µm 3 µJ −1 , with a spot size of 15 µm. Elemental analysis revealed 74.8% silicon, 16.6% carbon radicals, and 8.7% oxygen in the ablated material, suggesting the formation of SiO₂, which contributed to cracking in the heat‐affected zone. Surface roughness after cutting is as low as ≈28 nm. In situ plasma spectroscopy confirmed that SiC decomposed into Si+ networks and carbon radicals without undergoing melting. The ablation debris exhibited a layered structure, with carbon networks deposited over silicon, providing deeper insight into the material removal mechanism during laser machining. This approach, combining neural networks with physical simulation models, can be applied to any material or physical processing method.