1998
DOI: 10.1002/(sici)1096-9918(199803)26:3<195::aid-sia364>3.0.co;2-#
|View full text |Cite
|
Sign up to set email alerts
|

Resolution enhancement of x‐ray photoelectron spectra by maximum entropy deconvolution

Abstract: The maximum entropy method (MEM) is applied to the deconvolution of x‐ray photoelectron spectra. This method provides the least‐biased estimate of the unbroadened spectrum by using the Shannon information content as the regularizing functional. The large‐scale, non‐linear optimization problem is solved using a robust variable metric sequential‐quadratic programming (SQP) algorithm implemented on a personal computer (PC). The program is tested on simulated spectra and then shown to provide reliable resolution e… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
7
0

Year Published

2000
2000
2020
2020

Publication Types

Select...
6
3

Relationship

1
8

Authors

Journals

citations
Cited by 31 publications
(12 citation statements)
references
References 50 publications
0
7
0
Order By: Relevance
“…Furthermore, the noise of ARXPS data strongly affects the results. , The reconstruction is based on the assumption of model profiles but, due to the presence of noise, many models matching experimental data might exist. , For this reason, it is not possible to reconstruct the profiles by simply minimizing the weighted sum-of-square differences between calculated and measured data, especially if the surface layer is complex and consists of many components. However, the reconstruction of the depth profiles from ARXPS data is achievable by the maximum entropy method (MEM). , …”
Section: Introductionmentioning
confidence: 99%
“…Furthermore, the noise of ARXPS data strongly affects the results. , The reconstruction is based on the assumption of model profiles but, due to the presence of noise, many models matching experimental data might exist. , For this reason, it is not possible to reconstruct the profiles by simply minimizing the weighted sum-of-square differences between calculated and measured data, especially if the surface layer is complex and consists of many components. However, the reconstruction of the depth profiles from ARXPS data is achievable by the maximum entropy method (MEM). , …”
Section: Introductionmentioning
confidence: 99%
“…Analytical inversion leads to an amplification of instrumental noise and a highly oscillatory behavior of w tot ( n ) with possibly negative values, lacking physical meaning. , Sophisticated numerical approaches therefore have to be used for the calculation of w tot ( n ). Maximum entropy methods have successfully been applied in a number of related scientific problems in image reconstruction , spectroscopy and chromatography . We have therefore chosen this approach to calculate w tot by the procedure outlined in Figure (note the change in variable nomenclature, due to the fact that the procedure works with discrete data).…”
Section: Methodsmentioning
confidence: 99%
“…Combining both data sets to gain absolute molecular weight distributions free from band broadening requires a sophisticated data processing algorithm. Use was made of the so-called maximum entropy (MaxEnt) principle for this purpose, as data from the present kind of experiment contain a great amount of redundancy, usually leading to amplification of noise in the restored MWDs. With the employed system, MWDs of low and high polydispersity standards with a molecular weight of up to 15 kg mol −1 corrected for chromatographic band broadening could successfully be determined …”
Section: Introductionmentioning
confidence: 99%
“…The purpose of regularization is then to produce an estimate or reconstruction x of x from m, or, where the effects of blurring are negligible or of no interest, a reconstruction of x * b from m. One such regularization method is TPMEM. 9 It has been used to modify the MEM 4,12,13 by replacing the negative entropy term in the cost function with a two-point negative entropy term such that the cost function, F TM , is given by…”
Section: Theorymentioning
confidence: 99%