Influence of data processing methods on ATR-FTIR-based classification of pen inks

This research investigates the impact of data processing techniques on the accuracy and discrimination capability of ATR-FTIR spectroscopy in differentiating blue ballpoint pen inks through non-destructive analysis of ink directly from ink barrels. Six commercial pen brands (Artline, BiC, Faber-Castell, Faster, G’Soft, and Home & Office) were examined by analysing ink samples directly from the pen barrels without causing any damage to the pens. Initially, optimum parameters for establishing the ATR-FTIR method were identified to enhance data quality and reproducibility. Each functional group within the spectra was carefully identified and associated with specific ink components, such as solvents, colorants, resins/binders, and additives. This detailed interpretation was essential to understand the chemical composition of the inks before applying chemometric techniques for more efficient and objective data classification. By applying Principal Component Analysis (PCA) and Agglomerative Hierarchical Clustering (AHC), the study successfully discriminated between several pen samples based on their brand and type, demonstrating the potential of these chemometric techniques in forensic ink analysis. Through direct analysis of ink from the ink barrel surface, this study effectively eliminates paper-related interferences that could otherwise affect ink characterization. The establishment of optimum analytical parameters is essential to minimize variations during pen ink analysis and offers a reliable reference method for future studies. Furthermore, the application of PCA and AHC has proven to be effective statistical tools for chemical data classification, providing a more objective and efficient alternative to traditional manual comparison methods.