The line shape analysis of electron spectroscopy spectra by the artificial intelligence methods for identification of C sp2/sp3 bonds

Abstract

AbstractThe functionalized by the oxidation reaction multiwall carbon nanotubes, denoted as ox‐MWCNTs, and temperature modified are studied using the X‐ray induced Auger electron spectroscopy (XAES) and elastic peak electron spectroscopy reflection electron energy loss (EPES REELS). The artificial intelligence methods, like (i) the principal components (PC) and (ii) the k‐nearest neighbors (kNN) rule are applied for analysis of spectra line shapes. These methods apply spectra (i) transformation (PC) and (ii) n‐dimensional representation as vector points (kNN) and deal with distances between vector points. The spectra of temperature modified ox‐MWCNTs are identified on the basis of respective spectra recorded from the so‐called reference set, i.e., the spectra from graphite (C sp2 hybridizations) and diamond (C sp3 hybridizations). The kNN rule is more efficient than the PC for identification of carbon materials containing various C sp2 content and can be applied for quantification. The kNN rule indicated majority of C sp2 in temperature treated ox‐MWCNTs. The accuracy of the kNN rule is discussed.