atr-ir fingerprinting as a powerful method for identification of traditional medicine samples: a report of 20 herbal patterns

background and objectives: attenuated total reflectance-inferared (atr-ir) spectra can be used as a non-invasive fingerprinting approach in quality control of herbal samples. methods: twenty versatile herbal samples were subjected to attenuated total reflectance-inferared (atr-ir) spectroscopy followed by different clustering methods in order to determine by which method more reasonable classifications would be obtained. results: all classification methods (k-means, hca, pca and som) were able to discriminate the two medicinal seeds, hyocyamus niger and peganum harmala from other herbal samples. similarly, the starch samples were clustered in a reasonable method. in hca, one cluster included three types of starch samples: zea mays, oryza sativa and triticum aestivum. all the four classification methods were able to separate solanum tuberosum starch from other starch samples. hca and som, were able to classify leaf samples origanum vulgare and melissa officinalis belonging to lamiaceae family, in one category. crocus sativus and its adulterant carthamus tinctorius flowers were identified by pca, hca and som as different categories. conclusion: the result of this study can be utilized for identification and quality control of traditionally used medicinal plant samples in an unknown sample powder. such data could be the basis for preparing a data bank on iranian medicinal samples which in turn is used as a simple, fast and reliable method for characterization of herbal powders in pharmacopoeias.