Analysis and Classification of Liquid Samples Using Spatial Heterodyne Raman Spectroscopy

Gojani, Ardian B. and Palásti, Dávid J. and Paul, Andrea and Galbács, Gábor and Gornushkin, Igor B. (2019) Analysis and Classification of Liquid Samples Using Spatial Heterodyne Raman Spectroscopy. APPLIED SPECTROSCOPY, 73 (12). pp. 1409-1419. ISSN 0003-7028 (print); 1943-3530 (online)

Preview

Text Pre-proofSHSarticleAPS2019.pdf Download (932kB) | Preview

Abstract

Spatial heterodyne spectroscopy (SHS) is used for quantitative analysis and classification of liquid samples. SHS is a versionof a Michelson interferometer with no moving parts and with diffraction gratings in place of mirrors. The instrumentconverts frequency-resolved information into a spatially resolved one and records it in the form of interferograms.The back-extraction of spectral information is done by the fast Fourier transform. A SHS instrument is constructedwith the resolving power 5000 and spectral range 522–593 nm. Two original technical solutions are used as compared toprevious SHS instruments: the use of a high-frequency diode-pumped solid-state laser for excitation of Raman spectra anda microscope-based collection system. Raman spectra are excited at 532 nm at the repetition rate 80 kHz. Raman shiftsbetween 330 cm-1 and 1600 cm-1 are measured. A new application of SHS is demonstrated: for the first time, it is used forquantitative Raman analysis to determine concentrations of cyclohexane in isopropanol and glycerol in water. Two cali-bration strategies are employed: univariate based on the construction of a calibration plot and multivariate based on partialleast squares regression. The detection limits for both cyclohexane in isopropanol and glycerol in water are at a 0.5 mass%level. In addition to the Raman–SHS chemical analysis, classification of industrial oils (biodiesel, poly(1-decene), gasoline,heavy oil IFO380, polybutenes, and lubricant) is performed using the Raman–fluorescence spectra of the oils and principalcomponent analysis. The oils are easily discriminated showing distinct non-overlapping patterns in the principal componentspace.

Actions (login required)

Edit Item