Evaluation of Drude parameters for liquid Germanium nanoparticles through aerosol-based line-of-sight attenuation measurements. (March 2019)

• Record Type: Journal Article
• Authors: Daun, Kyle J.; Menser, Jan; Asif, Muhammad; Musikhin, Stanislav; Dreier, Thomas; Schulz, Christof
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 226(2019)
• Page Start: 146
• Page End: 156
• Publication Date: 2019-03
• Subjects: Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2019.01.021
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: Radiative properties of liquid Ge nanoparticles are measured between 205–700 nm. Drude parameters are inferred from extinction spectra via Bayesian inference. Inferred Drude parameters do not match values found with density and resistivity. Abstract: The objective of this study is to infer Drude model parameters for liquid germanium nanoparticles from extinction measurements made across an aerosol within a microwave plasma reactor using a halogen lamp (410–700 nm) and a laser-driven light source (205–585 nm). The plasma frequency and relaxation time are inferred using Rayleigh theory, Mie theory, and a fourth-order Mie approximation. These parameters are compared with those found using the ellipsometry-derived complex dielectric function as well as the bulk density and electrical resistivity of liquid germanium. The analysis is carried out in a probabilistic context using Bayesian inference, which accounts for both the measurement noise and model error. While all the candidate models can reproduce the shape of the experimentally-derived extinction spectra, the Bayesian inference showed that extinction-derived parameters differed from those obtained from the density and electrical resistivity in a statistically-significant way. This highlights the limitations of the free-electron model that underpins Drude theory, and suggests potential opportunities for model refinement.