Deconvolution method for obtaining directly the original in-depth distribution of composition from measured sputter depth profile. (August 2019)

• Record Type: Journal Article
• Authors: Lian, S.Y.; Wang, Z.J.; Wang, C.L.; Yan, X.L.; Lin, F.R.; Wang, J.Y.; Xu, C.K.
• Is Part Of: Vacuum. Volume 166(2019)
• Page Start: 196
• Page End: 200
• Publication Date: 2019-08
• Subjects: Depth profile -- Deconvolution -- MRI model -- Regularization -- TV-Tikhonov -- Noise; Vacuum -- Periodicals; 621.55
• DOI: 10.1016/j.vacuum.2019.05.008
• Languages: English
• ISSNs: 0042-207X

Abstract

Abstract: A new method is proposed to determine the original in-depth distribution of composition directly by the deconvolution of measured depth profiling data using the MRI (Mixing-Roughness-Information) depth resolution function. To overcome the ill-conditioned problem that is often experienced in the deconvolution procedure, the regularization method and the TV-Tikhonov algorithm are employed. The method is applied for the case that the original depth-concentration distribution is either a smooth function or a square wave function and the corresponding arithmetic mean errors are determined accordingly. The noise effects from "measured" data on the deconvoluted profile are also quantitatively evaluated. Highlights: A new method was proposed for deconvolution of measured depth profiling data. The stability of this method is tested for the cases of different noise levels and insufficient data point. It is demonstrated that this method is more powerful than the method in literature.