Influence of vacancies on indium atom distribution in InGaAs and InGaN compounds

It has been theoretically ascertained that for defect-free InGaAs and InGaN compounds the uniform distribution of indium atoms is more energetically preferable than the clustering distribution. The presence of gallium and arsenic vacancy in InGaAs and nitrogen vacancy in InGaN facilitates indium atom clustering distribution. It has been shown that the increase in the indium content in InGaAs and InGaN compounds leads to the decrease of the formation energy of gallium, arsenic and nitrogen vacancies.

ENERGETICS OF VHg-RELATED DEFECTS IN As-DOPED HgCdTe

Arsenic-doped HgCdTe usually exhibits compensated n-type conductivity, which is ambiguously attributed to isolated vacancies (V Hg ) or arsenic-vacancy complexes ( As Hg – V Hg and As Hg –2V Hg ). Our first-principles calculations clarified the correlation of these V Hg -related defects with the carrier compensation in As -doped HgCdTe by calculating the defect formation energies, as a function of atomic/electron chemical potentials. Under n-type condition, the lowest formation energy defect is found to be the As Hg donor followed by the V Hg acceptor, leading to a compensated n-type material. The arsenic-vacancy complexes are shallow acceptors but their high formation energies render them unlikely to be the compensating candidates. Their large binding energies, however, allow us to predict that the formation of the As Hg –2V Hg complex defect will be enhanced under postgrowth annealing treatment, in agreement with the arsenic activation model by Berding et al. [J. Electron. Mater.27, 605 (1998)].

Modeling of the Diffusion and Activation of Arsenic in Silicon Including Clustering and Precipitation

We have developed a diffusion and activation model for implanted arsenic in silicon. The model includes the dynamic formation of arsenic-vacancy complexes (As4V) as well as the precipitation of a SiAs phase. The latter is mandatory to correctly describe concentrations above solid solubility while the former are needed to describe the reduced electrical activity as well as the generation of self-interstitials during deactivation. In addition, the activation state after solid-phase epitaxy and the segregation at the interface to SiO2 are taken into account. After implementation using the Alagator language in the latest version of the Sentaurus Process Simulator of Synopsys, the parameters of the model were optimized using reported series of diffusion coefficients for temperatures between 700 °C and 1200 °C, and using several SIMS profiles covering annealing processes from spike to very long times with temperatures between 700 °C and 1050 °C and a wide distribution of implantation energies and doses. The model was validated using data from flash-assisted RTP and spike annealing of ultra-low energy arsenic implants.