Premium
Sputtering Yield Changes, Surface Movement and Apparent Profile Shifts in SIMS Depth Analyses of Silicon Using Oxygen Primary Ions
Author(s) -
Wittmaack K.
Publication year - 1996
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/(sici)1096-9918(199606)24:6<389::aid-sia135>3.0.co;2-l
Subject(s) - silicon , sputtering , secondary ion mass spectrometry , analytical chemistry (journal) , chemistry , oxygen , yield (engineering) , ion , boron , atomic physics , oxide , materials science , thin film , nanotechnology , metallurgy , physics , organic chemistry , chromatography
Depth profile analyses by secondary ion mass spectrometry (SIMS) are often carried out using oxygen bombardment at near‐normal beam incidence. During the initial stages of oxygen incorporation in the sample the secondary ion yield and the sputtering yield change significantly, notably in silicon, and stabilize only after a transition layer of thickness z tr has been eroded. In SIMS depth analyses these transient effects give rise to a depth scale offset z d U 0 and an associated profile shift z sh . A simple model is presented that provides means for determining z d U 0 and z tr from the difference between the initial and the final sputtering yield, the oxygen retention coefficient and the thickness of the oxide, w SiO2 , established under stationary bombardment of silicon. The results derived from the model are shown to be in good agreement with apparent shifts observed in depth profiling of boron implantation distributions, in which case z d U 0 ∽ z sh . The profile shift increases almost linearly with the oxygen energy E : z sh (nm)∽1.8 E (keV per O atom). The same number is derived from measurements on boron delta doping layers, provided that the centroid of the measured profile is used to quantify the shift. The transition depth z tr amounts to ∽2 z d U 0 and roughly equals the sum of the mean oxygen range 〈 z 〉 and range straggling σ, or ∽80% of the oxide thickness. This is much larger than the previously advocated value of z tr ⩽ 〈 z 〉 . With clean samples and at impact energies between 0.5 and 5 keV atom ‐1 the model predicts surface recession from the beginning of bombardment, but with a very small initial recession rate. The extrapolated zero‐fluence surface position or apparent crater offset z s U 0 is estimated to be negative (‘swelling’) but rather small: z s U 0 ∽‐0.15 w SiO2 . This is in accordance with most of the previously reported crater depth measurements, which suggested that the surface recedes directly proportional to the bombardment fluence with an undetectably small crater offset. With dirty or otherwise ill‐defined surfaces an initial swelling and a significantly enlarged negative crater offset might be observed.