Search

In this paper the authors discuss the issues surrounding the formation of ultrashallow n-type junctions in Ge. In general the n-type dopants are relatively difficult to activate and diffuse quickly, leading to high access resistances and limited capability to reduce the device dimensions, respectively. Sheet resistance (Rs) was calculated for n-type box-like profiles in Si and Ge. To achieve ITRS Roadmap targets in Si an active doping concentration of 1.5-2×1020 cm-3 is required, but due to the enhanced electron mobility it is predicted approximately 6×1019 cm-3 is sufficient in Ge.

All existing transistors are based on the use of semiconductor junctions formed by introducing dopant atoms into the semiconductor material. As the distance between junctions in modern devices drops below 10 nm, extraordinarily high doping concentration gradients become necessary. Because of the laws of diffusion and the statistical nature of the distribution of the doping atoms, such junctions represent an increasingly difficult fabrication challenge for the semiconductor industry. Here, we propose and demonstrate a new type of transistor in which there are no junctions and no doping concentration gradients. These devices have full CMOS functionality and are made using silicon nanowires. They have near-ideal subthreshold slope, extremely low leakage currents, and less degradation of mobility with gate voltage and temperature than classical transistors.

The authors investigate F as a non-dopant co-implant to suppress P and As diffusion in Ge. F was placed either overlaying the dopant profile, or deeper to counteract the influence of end-of-range defects, which is a common ultra-shallow junction optimization approach in Si substrates. It was determined that F outgasses extremely quickly from Ge, as a 1×1015cm−2 implanted dose escaped completely during a rapid thermal anneal as short as 1 sec, at 600 °C. This behavior is attributed to rapid diffusion, instability of F-defect clusters, and an aversion of F to reside substitutionally in the Ge lattice. The F diffusivity at 600 °C is in the order of some metal impurity diffusivities at that temperature.

Design and optimization of n-type doped regions in germanium by solid phase epitaxial recrystallization (SPER) have been studied by the authors. A systematic study is presented of process variables that influence activation and thermal stability, including preamorphization, coimplants, recrystallization temperature, and postrecrystallization thermal treatments. Unlike silicon, activation after recrystallization in germanium is not optimum where the postrecrystallization thermal budget is kept to a minimum. With the aid of modeling, a maximum peak activation of 7 X 10(19) cm(-3) was extracted. A steady increase in sheet resistance during postrecrystallization anneals confirms the formation of metastable activation by SPER. It is predicted that active concentrations of 6-8 X 10(19) cm(-3) are sufficient to meet targets for sub-20 nm technologies. (C) 2010 American Institute of Physics. (doi: 10.1063/1.3452345)