Dielectric characterisation of water and chloride solutions by THz spectroscopy: an effective two-state model

Read full text
Authors: Ermal Nikollari
Cited by: 0

Abstract

This work combines an experimentally robust THz–TDS extraction of water permittivity with a deliberately mesoscopic, minimally committal modeling strategy.On the experimental side, it reconstructs the complex permittivity in the THz band using a differential transmission scheme: two thicknesses of the same sample are compared (instead of using air as a reference), allowing the complex permittivity to be obtained in closed form while strongly suppressing common factors.On the modeling side, it interprets the measured permittivity as a coarse-grained observable and fits it with a minimal two-population picture: the macroscopic response is a weighted sum of two Debye-like relaxations, with the population weight constrained by independent IR measurements, plus an ultrafast exchange term between the two effective populations. The final pure-water fit yields two separated relaxation times (τ1≃8.36 ps and τ2≃0.46 ps) plus an ultrafast exchange time (τex≃0.042 ps), implying dwell times of order 80 fs in a two-state kinetic mapping. For monovalent chloride solutions, the analysis is limited to the THz window (no comparable broadband dataset), anchored to literature static permittivity and stabilized via additional constraints. Even so, the fits show consistent ion-specific trends, suggesting that ions primarily perturb the more structured/coherent fraction by shifting the equilibrium between the two effective populations and modifying their effective dynamics. Overall, the thesis contributes a differential THz–TDS protocol that simplifies and stabilises inversion, and a mesoscopic dielectric model (two populations + exchange) that provides a quantitative, interpretable broadband description—especially for pure water.

Fields of Study & Subjects

Terahertz technology and applications Material Dynamics and Properties Spectroscopy and Quantum Chemical Studies

Publication details

Journal / source
Figshare
Publication date
1 January 2026
Document type
Thesis