Premium
The issue of pseudoreplication when applying a statistical exploratory approach to extract relevant features from ToF‐SIMS spectra
Author(s) -
Dell'Anna Rossana,
Canteri Roberto,
Coppedè Nicola,
Iannotta Salvatore,
Bersani Massimo
Publication year - 2013
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.5251
Subject(s) - principal component analysis , univariate , multivariate statistics , deposition (geology) , spectral line , representation (politics) , secondary ion mass spectrometry , computer science , mass spectrum , set (abstract data type) , analytical chemistry (journal) , pattern recognition (psychology) , mass spectrometry , biological system , materials science , chemistry , artificial intelligence , machine learning , physics , chromatography , biology , paleontology , law , programming language , astronomy , sediment , politics , political science
The issue of automated peak selection in time‐of‐flight secondary ion mass spectrometry (ToF‐SIMS) spectra is examined in relation to the hierarchical nature of the experimental design and the related existence of observations which are not statistically independent. To avoid unreliable results, the presence of pseudoreplication should be taken into account correctly. A combination of the recommended univariate peak selection approach with some multivariate techniques, namely principal component analysis and heat map data representation, is proposed to highlight and analyze obtained results for a set of ToF‐SIMS spectra from copper phthalocyanine thin films grown on TiO 2 substrates by a supersonic beam deposition apparatus. New insight is obtained on the effectiveness of the deposition for different working parameters of the process. Copyright © 2013 John Wiley & Sons, Ltd.