High-wavenumber analysis of Raman and FTIR spectra of amino acids for biomedical applications

We describe the results of a systematic experimental and theoretical investigation of the high-wavenumber vibrational region (2400-3800 cm^-1) of amino acids using Raman and FTIR spectroscopy. While biomolecular vibrational analysis emphasizes traditionally the fingerprint region below 2000 cm^-1 , our results demonstrate that the high-wavenumber region alone contains specific structural informative characteristics to differentiate each amino-acid. We measured Raman and FTIR spectra of all twenty (20) proteinogenic amino acids in powder forms and computed their vibrational spectra using density-functional theory (DFT). For analysis, we focused on the high-wavenumber region (2400-3800 cm^-1) which appears to be dominated with C–H, N-H, and O-H stretching modes. The experimental spectra exhibit reproducible differences among the amino-acids with different peaks and lineshapes of the peaks, reflecting variations in the molecular structure of each amino-acid. We applied the principal component analysis (PCA) over the exclusive measured high-wavenumber region and observed clear separation among the amino acids, demonstrating that this range provides sufficient distinctive spectral information for molecular identification. The DFT calculations confirm this result. This insight from both the experimental data and the DFT calculations can be considered as the basis for the development of physicsreinforced classification models aimed at improving spectral classification, molecular recognition, and structural prediction of biomolecules with the potential use of the high-wavenumber bands only.