Rosetta/COSIMA: Principal Component Analysis Applied to Time-of-Flight Secondary Ion Mass Spectra
Abstract
Here we present results from an ongoing laboratory calibration campaign with the COSIMA RM. We have prepared cometary dust analogue samples (powder with grain size 10 μm) from natural and synthetic minerals (pyroxene, olivine, hydrous silicates, sulfide, etc.), which - except for hydrous minerals - have been identified in cometary matter [Burchell and Kearsley, 2009]. A large number of time-of-flight mass spectra of these samples was measured with the COSIMA RM. Spectra from the different mineral samples show characteristic mass lines, grossly reflecting the mineral composition. With Principal Component Analysis (PCA) we are able to separate the spectra of our mineral samples into individual groups. Thus, PCA - together with other chemometric techniques [Engrand et al., 2006, Varmuza et al., 2011] - is a useful tool for the classification of COSIMA mass spectra. Further experiments have to show whether it can also be used for the classification of organic compounds and mixtures of, e.g., organic compounds and mineral grains.