Hanyang University researchers gain control over where semiconductor crystals first form - EurekAlert!

Controlling nucleation sites shifts epitaxial growth from statistical management to deterministic engineering. This is not a process tweak—it is a root-level intervention in defect propagation and interfacial strain. Whoever dictates where the first atom lands controls the entire defect budget of the epitaxial stack. Technical cascade: The breakthrough strikes the core bottleneck of heteroepitaxy. In GaN/SiC/Ga₂O₃ power devices, substrate-epitaxy interface defects still destroy 15-25% of yield. Upstream, AMAT, Lam, and TEL face reactor geometry redesign pressure. Downstream, 200mm SiC wafer yield could gain 3-5 points within 18 months, directly rewriting the power semiconductor cost curve. Market dynamics: Korea already commands 40%+ of global DRAM and 60%+ of NAND. University-level foundational IP strengthens the Korean stack thesis. Samsung and SK will likely pilot-integrate within 12 months. TSMC (Taiwan, China) and Intel will accelerate in-house nucleation control R&D to avoid losing process leverage in compound semiconductors. The real disruption sits in the $30B compound semiconductor materials market—SiC and GaN EV supply chain timelines could compress by 12-18 months. Risk and long tail: If patent claims cover reactor architecture, IP friction with US equipment giants is inevitable. Within 24 months, expect 3-5 patent families, 2-3 licensing deals, and a structural inflection in SiC power device cost curves.