Space-resolved Chemical Information from Infrared Extinction Spectra

TL;DR

This paper introduces a novel method based on the Lorentz model to extract pure chemical absorbance spectra from infrared extinction data, effectively removing scattering effects and spectral distortions without needing reference spectra.

Contribution

The method automatically satisfies Kramers-Kronig relations and determines particle sizes and layered structures, advancing spectral analysis of dielectric spheres.

Findings

Accurately extracts pure absorbance spectra from simulated data.

Effectively removes scattering effects from experimental spectra.

Works reliably even with uncorrected raw spectra.

Abstract

Based on the classical Lorentz model of the index of refraction, a new method is presented for the extraction of the complex index of refraction from the extinction efficiency of homogeneous and layered dielectric spheres that simultaneously removes scattering effects and corrects measured extinction spectra for baseline shifts, tilts, curvature, and scaling. No reference spectrum is required and the method automatically satisfies the Kramers-Kronig relations. Thus, the method yields pure absorbance spectra for unambiguous interpretation of the chemical information of the sample. In the case of homogeneous spheres, the method also determines the radius of the sphere. In the case of layered spheres, the method determines the radii of the layers and the substances within each layer. Only a single-element detector is required. Using simulated $Q_{ext}$ data of polymethyl-methacrylate (PMMA) and polystyrene homogeneous and layered spheres we show that our reconstruction algorithm is reliable and accurately extracts pure absorbance spectra from $Q_{ext}$ data. Reconstructing the pure absorbance spectrum from a published, experimentally measured raw absorbance spectrum shows that our method also works for uncorrected spectra, reliably removes scattering effects and, given shape information, simultaneously corrects for spectral distortions.