Dissolution of dopant-vacancy clusters in heavily doped semiconductors via millisecond annealing

The aggressive reduction of the channel length in transistors needs the high doping of the channel region, while the contact area requires doping beyond 1020 cm-3 to ensure low-resistance ohmic contacts. Similar problems apply to wide-band gap semiconductors where the p-type doping is challenging, mainly due to the high activation energy for acceptors, low equilibrium solid solubility and deactivation of acceptors by the formation of acceptor-vacancy clusters. Recently, we have shown that ultra-doped n-type and p-type Ge with a carrier concentration above 1020 cm-3 can be achieved by applying non-equilibrium methods like ion implantation followed by millisecond-range flash-lamp annealing (FLA) [1-3]. The n-type doping of Ge is a self-limiting process due to the formation of vacancy-donor complexes (DnV with n ≤ 4) that deactivates the donors [4]. Based on data density functional theory calculations, at temperature higher than 850 K, the concentration of D4V clusters progressively decreases liberating unbounded vacancies and donor atoms. The same effect is observed for p-type Ge and in III-V semiconductors. Here, we report on experiments and theoretical calculations solving the basic problem of donors and acceptors deactivation in heavily doped semiconductors. The dissolution of donor/acceptor-vacancy clusters in heavily doped semiconductors is achieved by ms-range FLA with a peak temperature close to the melting point of the semiconductor. Positron annihilation lifetime spectroscopy (PALS) reveals that dopant-vacancy clusters are the main defect centers deactivating both acceptors and donors. Millisecond-range high-temperature treatment dissociates the dopant-V clusters and, as shown by SIMS, fully suppresses the dopant diffusion in both group IV semiconductors and III-V compound semiconductors. For the first time, using structural characterization (PALS, SIMS) and electrochemical capacitance-voltage profiling combined with DFT calculations, we were able to address, understand, and solve the fundamental problem hindering the doping of semiconductors above the solid solubility limit.