Time-of-flight secondary ion mass spectrometry of a range of coal samples: A chemometrics (PCA, cluster, and PLS) analysis

This paper documents time-of-flight secondary ion mass spectrometry (ToF-SIMS) analyses of 34 different coal samples. In many cases, the inorganic Na+, Al+, Si+, and K+ ions dominate the spectra, eclipsing the organic peaks. A scores plot of principal component 1 (PC 1) versus principal component 2 (PC2) in a principal components analysis (PCA) effectively separates the coal spectra into a triangular pattern, where the different vertices of this pattern come from (i) spectra that have a strong inorganic signature that is dominated by Na+, (ii) spectra that have a strong inorganic signature that is dominated by Al+, Si+, and K+, and (iii) spectra that have a strong organic signature. Loadings plots of PCI and PC2 confirm these observations. The spectra with the more prominent inorganic signatures come from samples with higher ash contents. Cluster analysis with the K-means algorithm was also applied to the data. The progressive clustering revealed in the dendrogram correlates extremely well with the clustering of the data points found in the scores plot of PC1 versus PC2 from the PCA. In addition, this clustering often correlates with properties of the coal samples, as measured by traditional analyses. Partial least-squares (PLS), which included the use of interval PLS and a genetic algorithm for variable selection, shows a good correlation between ToF-SIMS spectra and some of the properties measured by traditional means. Thus, ToF-SIMS appears to be a promising technique for the analysis of this important fuel.