AVS 58th Annual International Symposium and Exhibition
    Applied Surface Science Division Tuesday Sessions
       Session AS-TuP

Paper AS-TuP21
Comparison between the Continuous and Discrete Model to Assess the Thickness of SiO2 Layers on Si with XPS Data

Tuesday, November 1, 2011, 6:00 pm, Room East Exhibit Hall

Session: Applied Surface Science Poster Session
Presenter: Milton Vazquez-Lepe, Cinvestav Queretaro Mexico
Authors: M.O. Vazquez-Lepe, Cinvestav Queretaro Mexico
P.G. Mani-Gonzalez, Cinvestav Queretaro Mexico
A. Mendoza-Galvan, Cinvestav Queretaro Mexico
A. Herrera-Gomez, Cinvestav Queretaro Mexico
Correspondent: Click to Email
X-ray Photoelectron Spectroscopy (XPS) is a valuable tool for assessing the thickness of surface layers in the 0 to 8 nm range. The prediction of a core-level XPS signal from a solid requires assuming a specific volumetric distribution of such core levels. Since the core levels are spaced close to the atomic nucleus, a possible approximation is to model their volumetric distribution as a sum of delta functions at the atomic sites. Another approach is to consider that the source of the photoelectrons is uniformly distributed within the volume of the solid. The latter is called the continuous approximation (CA) and the former the discrete approximation (DA). The CA is widely employed; it leads to an expression for the XPS signal proportional to cλ sin α, where c is the volumetric ddensity, λ is the effective attenuation length, and α is the take off angle. The DA applied to crystalline structures leads to an expression proportional to s[1-exp(-a/(λ sin α))]-1 exp(-a/(2λ sin α)), where a is the spacing of the atomic planes parallel to the surface, and s is the atomic plane surface concentration (s/a = c) [1]. When a/λ<<1, the DA expression reduces to the corresponding to CA. Applying the DA to amorphous materials requires defining an “effective” or “average” atomic spacing (ae) and surface concentration (se). In this paper we compare the differences on the thickness of silica layers on Si[001] calculated with these two approaches. The analysis was done by employing XPS data at one angle or at various angles (ARXPS). The samples employed were thermally grown silica at 800 °C for 16 s, 75 s, 145 s, 212 s and 7 min. The results are also compared to more standard methods [2], such as that employing the parameter R (ratio of the Si 2p XPS intensities for clean Si and thick SiO2). The thickness of the silica layers were also assessed from Transmission Electron Microscopy (TEM) images and correlated to Ellipsometry measurements. It was found that the approach employing the DA and various angles better reproduced the TEM measurements.

[1] “Self consistent ARXPS analysis for multilayer conformal films with abrupt interfaces.” A. Herrera-Gomez. Internal Report 2007 (http://www.qro.cinvestav.mx/~aanalyzer/arxpsAnalysisSharpIntefaces.pdf).

[2] Cumpson PJ. Surf. Interface Anal. 2000; 29: 403.