Premium
Ultrathin SiO 2 on Si. VI. Evaluation of uncertainties in thickness measurement using XPS
Author(s) -
Seah M. P.
Publication year - 2005
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.2020
Subject(s) - x ray photoelectron spectroscopy , materials science , analytical chemistry (journal) , chemistry , chemical engineering , environmental chemistry , engineering
Measurements of the thicknesses of SiO 2 on Si in the range 1.5–8 nm can be made by XPS with precisions as high as 0.025 nm at one standard deviation, but in the past this has been overshadowed by systematic uncertainties of the order of 20% of the thickness arising from the uncertainty in the relevant electron attenuation length. This is shown no longer to be the case. An analysis is presented, of the contributions to the measurement precision arising from the counting statistics, peak fitting and setting of the emission angle for the detected electrons. With traditional commercial equipment, the angle‐setting accuracy of 0.5° can be the limiting factor. For comparison between laboratories, further terms degrade the comparability: a typical 1% from the consistency of any samples used for the comparison; a more significant term from the angular acceptance of those spectrometers that accept more than ±10° of the emission angles; contributions from the variability of the procedures for analysing the peak intensities for the quantification; and variations in the algorithms used. For absolute quantification in the above thickness range, a recent calibration of the attenuation length enables the uncertainty to be kept to ∼2% at 95% confidence but the above variability in procedures in general analytical laboratories leads to this value degrading often to as poor as 0.4 nm, again at 95% confidence. Copyright © 2005 John Wiley & Sons, Ltd.