Nondestructive Identification of Black Pen Inks Using Surface-enhanced Raman Spectroscopy and Chemometrics

To investigate the feasibility of surface-enhanced Raman spectroscopy (SERS) combined with chemometric methods for the nondestructive and rapid identification of black pen inks.

By optimizing silver nanoparticle substrates and spectral acquisition parameters, SERS spectra of 30 black pen inks were collected. Classification models were constructed using K-means clustering algorithm (K-Means), K-nearest neighbor classification algorithm (KNN), and random forest (RF) algorithm, with their performance evaluated based on accuracy and F1-score.

SERS effectively suppressed fluorescence background and enhanced Raman signals, yielding distinct characteristic peaks for all tested samples. The carbon material D band and G band located at 1 338~1 392 cm^-1 and 1 583~1 619 cm^-1 were also identified. The KNN demonstrated optimal classification performance, as its results were fully consistent with those obtained from direct SERS discrimination.

The combination of SERS technology and chemometric methods can be applied to the accurate classification of black pen inks, providing a nondestructive, rapid and reliable approach for forensic questioned document examination.