Analysis of TOF-SIMS spectra by machine learning

Abstract

Time-of-flight secondary ion mass spectrometry (TOF-SIMS), which detects chemical information of a solid sample at the most upper surface less than 2 nm and chemical images with a high spatial resolution approximately 100 nm, is widely applied to analyze biological samples, organic materials, and electronic materials. Because the TOF-SIMS spectra and image data are generally very complicated, numerical analysis methods are often required to interpret TOF-SIMS data. Multivariate analysis such as principal component analysis has successfully been applied to TOF-SIMS data interpretation. Recently, sparse modeling and machine learning, which are now applied to many fields, are also employed for TOF-SIMS data analysis. In this article one of the latest examples of machine learning application to TOF-SIMS data is introduced.