Chemometric Analysis of Chromatographic Data

A short description of the tools used can be foimd in the Appendix. For more details on the background of chemometrics/multivariate analysis, one should refer to the additional literature. First of all, a data analysis is briefly considered, which was kindly performed by Prof. Dr. Heinz Zwanziger, Merseburg, Germany. 

Tree Diagram for 24 Variables (without retention) Objects 30 Columns 1-Pearson r Ward s method 

Even if this analysis only takes the data of one experiment into account ( hydro-phobicity ), the restdts obtained show the expressiveness of chemometric tools. 

The clustering of 24 variables of 30 columns is depicted in Fig. 32. Two clusters can be discerned. The first cluster ( a ) includes k values, values for the specific surface area, and the carbon content. The other variables, e.g. the asymmetry factor (AS), the plate number (N), and the diameter of the pores are in the second cluster ( b ). 

Cluster analysis of 30 columns further led to the following results  

---