Kinetically controlled dewetting of thin GaAs cap from an ErAs/GaAs nanoparticle composite layer

JOURNAL OF VACUUM SCIENCE & TECHNOLOGY B(2018)

引用 3|浏览10
暂无评分
摘要
ErAs metal nanoparticles (NPs) embedded in GaAs have multiple applications in plasmonic, terahertz, and tunneling devices. Growing a high quality thin GaAs layer over the ErAs NP layer is vital to these applications. In this work, the authors study the surface stability of a thin GaAs cap (1-5 nm) annealed in a temperature range of 450-620 degrees C. The thin GaAs cap covered a single layer of ErAs NPs [0.5-1.33 monolayer (ML)] grown using molecular beam epitaxy on GaAs(001) substrates at 450-500 degrees C. For 1.33 ML ErAs coverage, although a 1 nm GaAs cap exhibited a root-mean-square surface roughness close to 0.3 nm, the authors expected that 1 nm GaAs was not thick enough to overgrow the NPs in a height of 3-4 nm; thus, a large number of pinholes should be left on the surface. By increasing the GaAs cap thickness to 3 nm, the authors were able to achieve atomically smooth surfaces with few remaining pinholes. At a lower coverage of ErAs, 0.5 ML, the authors were able to achieve atomically flat pinhole-free GaAs caps with a thickness of 3-5 nm. However, the key finding from this study is that distinct changes in the surface morphology occurred upon annealing depending on the film thickness and NP density. In the case of 1 nm GaAs caps, clumps were formed when annealed, whereas in the case of GaAs caps of 3-5 nm in thickness, the GaAs film uncharacteristically dewetted at the ErAs NP/GaAs composite interface. Thermodynamically, this dewetting is driven by the high interfacial energy resulting from the difference in the crystal structure between GaAs and ErAs (zinc-blende and rock salt); however, surface mobility plays an important kinetic role in this process. It has been demonstrated that the dewetting can be prevented by combining a higher As overpressure, a low growth/annealing temperature, lower surface coverage of ErAs NP, and thicker GaAs caps.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要