IR spectroscopy is widely used in identification and characterization of both organic and inorganic matters. IR spectra is unique in terms of their complexity of the information that are posed in spectra. A large number of peaks appear on the spectra. Additionally, the peaks largely vary in absorption cross-sections and broadness. Along those lines, we are interested in understanding how various naturally occurring and man-made matters exhibits common and individually unique spectral features. The goal was to develop a computational approach that allows for rapidly categorizing a given sample into pre-defined classes/subclasses based on the characteristics that can be observed in IR spectra. For this purpose, we have developed a software that performs cluster analysis of IR spectra. Our observation was that the K-means clustering could be applied in finding the groups of naturally occurring and man-made matters that show similar spectral features.

Computational study on structure of IR spectral library using cluster analysis

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.