Premium
Depth profiling of shallow arsenic implants in silicon using SIMS
Author(s) -
Clegg J. B.
Publication year - 1987
Publication title -
surface and interface analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.52
H-Index - 90
eISSN - 1096-9918
pISSN - 0142-2421
DOI - 10.1002/sia.740100704
Subject(s) - sputtering , silicon , analytical chemistry (journal) , materials science , range (aeronautics) , rutherford backscattering spectrometry , oxygen , yield (engineering) , chemistry , atomic physics , thin film , optoelectronics , nanotechnology , physics , organic chemistry , chromatography , metallurgy , composite material
The analytical conditions necessary for the accurate measurement of shallow As implant profiles in crystalline Si have been critically evaluated. It is shown that near‐normal oxygen bombardment (θ = 2°) gives rise to a drive‐in of the As peak toegether with an enlarged exponential tail due to segregation processes associated with SiO 2 formation in the near‐surface region. At θ = 45°, both effects are greatly reduced. Thus, for enhanced depth resolution, it is essential to use oblique bombardment with low probe energies. Silicon sputter yield variations in the pre‐equilibrium region are shown to be much smaller with θ = 45° and the differential peak shift amounts to only 0.6 nm per keV O + compared with 3.0 nm at θ = 2°. The oblique sputter conditions do, however, lead to an increase in the thickness ( w ) of the pre‐equilibrium region together with a loss in analytical sensitvity. The use of an oxygen jet is shown to reduce w to about 1 nm but gives rise to ‘chemically induced’ profile tailing with low energy (5 keV) As implants. Systematic measurements with different probe energies (1 and 2 keV O 2 + ) show that As implants down to 10 keV are reliably profiled with θ = 45°. Comparison of our As range data with published Rutherford backscattering spectrometry values shows good agreement for the standard deviation while the mean projected range shows a systematic difference of ≃ 10%.