WuB122009.pdf (4.86 MB)
InAs-on-SOI MOSFETs with Extreme Lattice Mismatch
thesis
posted on 2009-12-14, 00:00 authored by Bin WuBoth molecular beam epitaxy (MBE) and metal organic chemical vapor deposition (MOCVD) have been used to explore the growth of InAs on Si. Despite 11.6% lattice mismatch, planar InAs structures have been observed by scanning electron microscopy (SEM) when nucleating using MBE on patterned submicron Si-on-insulator (SOI) islands. Planar structures of size as large as 500 ÌÄ' 500 nm2 and lines of width 200 nm and length a few microns have been observed. MOCVD growth of InAs also generates single grain structures on Si islands when the size is reduced to 100 ÌÄ' 100 nm2. By choosing SOI as the growth template, selective growth is enabled by MOCVD. Post-growth pattern-then-anneal process, in which MOCVD InAs is deposited onto unpatterned SOI followed with patterning and annealing of InAs-on-Si structure, is found to change the relative lattice parameters of encapsulated 17/5 nm InAs/Si island. Observed from transmission electron diffraction (TED) patterns, the lattice mismatch of 17/5 nm InAs/Si island reduces from 11.2 to 4.2% after being annealed at 800 Ìâå¡C for 30 minutes. High-k Al2O3 dielectrics have been deposited by both electron-beam-enabled physical vapor deposition (PVD) and atomic layer deposition (ALD). Films from both techniques show leakage currents on the order of 10-9A/cm2, at ~1 MV/cm electric field, breakdown field > ~6 MV/cm, and dielectric constant > 6, comparable to those of reported ALD prior arts by Groner. The first MOSFETs with extreme lattice mismatch InAs-on-SOI channels using PVD Al2O3 as the gate dielectric are characterized. Channel recess was used to improve the gate control of the drain current.
History
Date Modified
2017-06-02Defense Date
2009-12-01Research Director(s)
Patrick FayCommittee Members
Patrick Fay Thomas Kosel Gregory SniderDegree
- Doctor of Philosophy
Degree Level
- Doctoral Dissertation
Language
- English
Alternate Identifier
etd-12142009-145002Publisher
University of Notre DameProgram Name
- Electrical Engineering
Usage metrics
Categories
No categories selectedLicence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC