Application of chemometric tools for ion analysis in time-of-flight secondary mass spectrometry and ion mobility spectrometry

High resolution time-of-flight secondary ion mass spectrometry (HR TOF-SIMS) is a powerful surface analytical method. For complex samples, this technique may yield intricate spectra that are difficult to interpret visually. Chemometric methods are useful for data analysis. However, these methods require that spectra are represented in a matrix format. Variances in mass measurements caused by calibration or instrumental effects may present difficulties in properly aligning mass spectral peaks into the correct columns of the data matrix. Cluster analysis of resolution elements is proposed as an alternative approach to construct the data matrix. An automated method for optimizing the data alignment is presented and evaluated for standard steel samples.; Ion mobility spectrometry (IMS) has a limited linear range. Nonlinear calibration methods, such as neural networks are ideally suited for IMS due to their capability of modeling complex systems. Many neural networks suffer from long training times and overfitting. Cascade correlation neural networks (CCN) are interesting, because they train at fast rates and automatically configure their own topology. By using the decay parameter in training neural networks, reproducible and general models may be obtained at the cost of longer training times. CCN networks were trained to furnish both quantitative and qualitative prediction for a complex IMS data set. The advantage of rapid training is that replicate neural networks may be obtained. Partial least-squares regression (PLS) is used as a comparative method. The CCN with decay rate one order of magnitude larger than the learning rate, achieves significantly better results than those obtained from an optimal PLS model.; Deconvolution is a powerful method to improve spectral resolution. A Fourier deconvolution (FD) algorithm has been developed that retains quantitative spectral information and incorporates automated filtering. The advantages of the algorithm include simplicity, fast speed, and reliability. In addition, an automatic cutoff in the frequency domain is calculated and used for filtering high-frequency noise from the IMS spectra. The deconvolved IMS spectra were evaluated by principal component analysis (PCA). Evaluation of principal component scores demonstrates that deconvolution can improve the separation of spectra from different concentrations. In addition, PCA is useful for optimization of deconvolution.