authentication of ink samples with atr-ftir spectroscopy associated with pls-da method

Recently, adulteration and abuse in the production of industrial products with the aim of gaining more profit have caused the increase of consumers demand for supervision and authentication of these products. one of the strategic products used in different fields is ink. on the other hand, vibrational spectroscopic methods as a fast, strong and non-destructive tool provide useful and effective information on the chemical structure of the ink such as solvents and resins [1-3]. in this study, infrared (ir) spectroscopy and chemometrics statistical method were used to classify seven different brands of pen ink. available samples were classified into 7 classes (5 original and 2 adulterated). then, preprocessing methods were performed using mean centering (mc) and standard normal variate (snv) method. in the next step, they were divided into two sets of 44 samples as training set and 19 samples as testing set by the kennard-stone algorithm, being analyzed by partial least squares discriminant analysis (pls-da) method. error rate of these models with genetic algorithm (ga) and without it were 8% and 13% for the pls-da, respectively. thus, it can be concluded that the presented method demonstrated the rapid, simple, without any chemical preparation and effectiveness classification method of ink samples by ir spectroscopy combined chemometric technique