Investigation of Boron Transient Diffusion in Sub-micron Patterned Silicon by Scanning Capacitance Microscopy

碩士 === 國立中央大學 === 物理研究所 === 99 === Current microelectronics chip can be composed of thousands of microarrays that contain up to millions of physically identical transistors layout in vastly different micro-environment. Systematic threshold voltage (Vth) variation due to the detailed difference in th...

Full description

Bibliographic Details
Main Authors: Fei-bai CHEN, 陳飛白
Other Authors: Wei-yen WOON
Format: Others
Language:en_US
Published: 2011
Online Access:http://ndltd.ncl.edu.tw/handle/07900423829794430110
Description
Summary:碩士 === 國立中央大學 === 物理研究所 === 99 === Current microelectronics chip can be composed of thousands of microarrays that contain up to millions of physically identical transistors layout in vastly different micro-environment. Systematic threshold voltage (Vth) variation due to the detailed difference in the microenvironment has been shown in many electrical assessments. In this work, we have designed an experimental platform for investigating the dependence of dimensionality in two dimensional boron diffusion lengths (Ldi f f ). We systematically vary the ion implantation window length scales in both length (l) and width (w) directions using photolithography process. The two dimensional Ldi f f are measured with plane view scanning capacitance microscopy (SCM). The Ldi f f in width shrunk patterns exhibit stronger diffusion, especially in ion implantation windows with larger l, namely, boron transient diffusion roll-off. This observation suggest there is effectively more interstitial (Is) sources within the proximity of B-Is interaction range during annealing and lead to more significant transient enhanced diffusion (TED) at larger confinements. The normalized Ldi f f for ion implantation boundaries length scales ranging from 0.3 micron to 5 micron shows five folds difference. The normalized curves for both categories of patterns overlap, indicating similar physical mechanism in play for the two cases. We have developed a non-linear logistics model. We can successfully fit the experimental data with the above model by considering only the difference in dimensionality. In particular, we found a 3/5 ratio for the linear growth coefficients of effective Is supersaturation with respect to the ion implantation boundary dimensions between the two patterns. We relate this coefficient ratio to number of interstitial injection boundaries available within B-Is interaction range.