1. Optical characterization limits of nanoparticle aggregates at different wavelengths using approximate Bayesian computation
- Author
-
Hakan Erturk and Ozan Burak Eriçok
- Subjects
Effective radius ,Radiation ,Materials science ,010504 meteorology & atmospheric sciences ,Discrete dipole approximation ,Inverse problem ,01 natural sciences ,Atomic and Molecular Physics, and Optics ,Light scattering ,Computational physics ,Characterization (materials science) ,010309 optics ,Wavelength ,0103 physical sciences ,Particle-size distribution ,Limit (mathematics) ,Spectroscopy ,0105 earth and related environmental sciences - Abstract
Optical characterization of nanoparticle aggregates is a complex inverse problem that can be solved by deterministic or statistical methods. Previous studies showed that there exists a different lower size limit of reliable characterization, corresponding to the wavelength of light source used. In this study, these characterization limits are determined considering a light source wavelength range changing from ultraviolet to near infrared (266–1064 nm) relying on numerical light scattering experiments. Two different measurement ensembles are considered. Collection of well separated aggregates made up of same sized particles and that of having particle size distribution. Filippov's cluster-cluster algorithm is used to generate the aggregates and the light scattering behavior is calculated by discrete dipole approximation. A likelihood-free Approximate Bayesian Computation, relying on Adaptive Population Monte Carlo method, is used for characterization. It is found that when the wavelength range of 266–1064 nm is used, successful characterization limit changes from 21–62 nm effective radius for monodisperse and polydisperse soot aggregates.
- Published
- 2018
- Full Text
- View/download PDF