Vacancy-engineering-induced dislocation inclination in III-nitrides on Si substrates

Jie Zhang, Xuelin Yang, Yuxia Feng, Yue Li, Maojun Wang, Jianfei Shen, Lai Wei, Danshuo Liu, Shan Wu, Zidong Cai, Fujun Xu, Xinqiang Wang, Weikun Ge, Bo Shen

Research output: Contribution to journalArticlepeer-review

27 Citations (Scopus)

Abstract

The incorporation of point defects into semiconductors could substantially tailor their optical and electrical properties as well as the spin-based quantum properties. In terms of structural properties, however, efforts have seldom been devoted to the relevant aspects. Herein, we propose point defects engineering by intentionally introduced vacancies to improve the structural properties. GaN-on-Si are selected as a paradigm to demonstrate the applicability of this approach. By tuning the growth stoichiometry, nonequilibrium Ga vacancies are intentionally introduced and absorbed by dislocation cores, which leads to dislocation inclination and annihilation in GaN. In addition, this dislocation inclination can proceed without relaxing the compressive lattice stress. These together enable high quality GaN thick layers on Si substrates with dislocation density of 1.6×108cm-2 and a record electron mobility of 1090cm2/Vs at a carrier density of 1.3×1016cm-3. With these advances, a quasivertical GaN Schottky barrier diode with the lowest specific on-resistance of 0.95mω/cm2 and highest on/off ratio of 1010 on Si substrates is demonstrated. These results demonstrate the promise of point defect engineering as a strategy to improve the structural properties and pave the way for high-performance III-nitride based electronic and optoelectronic devices on Si platforms.

Original languageEnglish
Article number073402
JournalPhysical Review Materials
Volume4
Issue number7
DOIs
Publication statusPublished - Jul 2020
Externally publishedYes

Fingerprint

Dive into the research topics of 'Vacancy-engineering-induced dislocation inclination in III-nitrides on Si substrates'. Together they form a unique fingerprint.

Cite this