Mechanisms of arsenic clustering in silicon

TL;DR

This paper proposes a model for arsenic clustering in silicon emphasizing negatively charged complexes, with calculations matching experimental data and improving understanding of high concentration arsenic diffusion.

Contribution

The paper introduces a novel model focusing on doubly negatively charged arsenic clusters, aligning well with experimental results and enhancing arsenic diffusion simulations.

Findings

Doubly negatively charged clusters with two arsenic atoms fit experimental data.

Electron density peaks and then decreases with increasing arsenic concentration.

The model improves simulation accuracy of high concentration arsenic diffusion.

Abstract

A model of arsenic clustering in silicon is proposed and analyzed. The main feature of the proposed model is the assumption that negatively charged arsenic complexes play a dominant role in the clustering process. To confirm this assumption, electron density and concentration of impurity atoms incorporated into the clusters are calculated as functions of the total arsenic concentration. A number of the negatively charged clusters incorporating a point defect and one or more arsenic atoms are investigated. It is shown that for the doubly negatively charged clusters or for clusters incorporating more than one arsenic atom the electron density reaches a maximum value and then monotonically and slowly decreases as total arsenic concentration increases. In the case of doubly negatively charged cluster incorporating two arsenic atoms, the calculated electron density agrees well with the experimental data. Agreement with the experiment confirms the conclusion that two arsenic atoms participate in the cluster formation. Among all present models, the proposed model of clustering by formation of doubly negatively charged cluster incorporating two arsenic atoms gives the best fit to the experimental data and can be used in simulation of high concentration arsenic diffusion.