Back to Search Start Over

Si/InP direct wafer bonding: A first-principles study.

Authors :
Duan, Xueyi
Kang, Ruyan
Liu, Zehan
Zuo, Zhiyuan
Zhao, Jia
Source :
Computational Materials Science. May2024, Vol. 241, pN.PAG-N.PAG. 1p.
Publication Year :
2024

Abstract

[Display omitted] • Innovatively employed first-principles calculations to analyze the atomic-level bonding principles of heterogeneous wafer bonding interfaces. • Combines molecular dynamics and reaction kinetics for a multidisciplinary approach. • Presents eight interface models to explore bonding mechanisms. • Investigated annealing and oxygen plasma activation effects for optimizing bonding processes in optoelectronic integration. The direct bonding mechanism of Si/InP is systematically investigated using first-principles calculations. Charge density analysis reveals that Si can be bonded with either In or P and is polar covalent. The single Si-P bond is stronger than the Si-In bond and has a shorter bond length, the Si-P bond length is 2.356 Å compared to 2.590 Å for the Si-In ones, as it involves the sharing of electron pairs and a large transfer and redistribution of electrons. Notably, molecular dynamics simulations show that the interfacial strength of Si bonded to In is greater at room temperature. This is because, after sufficient relaxation, although the average strength of the Si-P bonds is still greater than that of Si-In, the number of Si-In bonds is approximately four times that of the Si-P bonds. The radial distribution function of the atoms in the interfacial layer further confirms that the terminal interface is stronger than the P-terminal interface. Appropriate annealing treatment can effectively increase the number and strength of bonds at the interface. Finally, the bonding mechanism of the oxygen plasma activation is studied using reaction kinetics. These findings contribute to the precise control and optimization of bonding processes, ultimately improving the reliability of optoelectronic integration. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
09270256
Volume :
241
Database :
Academic Search Index
Journal :
Computational Materials Science
Publication Type :
Academic Journal
Accession number :
177063690
Full Text :
https://doi.org/10.1016/j.commatsci.2024.113061