Precision radio-frequency and microwave dielectric spectroscopy and characterization of ionic aqueous solutions
Excessive amounts of chemicals and ions flowing into water sources, which are mainly due to efflux from agricultural lands, cause serious environmental and human-health related concerns. The lack of affordable and real-time monitoring systems for these contaminants limits effective conservation and management strategies. To establish a basis for developing an effective, fast, real-time, and affordable sensing system, dielectric spectroscopy has been applied to characterize agriculturally-relevant aqueous solutions of most commonly found ions in tile drainage water. Dielectric spectra of aqueous sodium chloride (NaCl), sodium nitrate (NaNO3), and sodium sulphate (Na2SO4) ionic solutions, which are the common pollutants found in agricultural tile drainage waters in Iowa and the United States, were measured over the frequency range from 200 MHz to 20 GHz, at temperatures 5 Ã Â°C to 30 Ã Â°C in 5 Ã Â°C increments and for concentrations in the range 0 to 20 millimoles per liter.
In this work, the measured dielectric spectra were fitted with a Debye model using a non-linear, weighted, least-squares analysis. Uncertainties due to random and systematic errors, that contribute to the measured dielectric spectra and become critical in the context of low concentration aqueous solutions, have been assessed. Moreover, two methods of calculating associated uncertainties of the fitting parameters, covariance matrix and Monte Carlo methods have been performed. The results show that the numerical approach taken by the Monte Carlo method, while yielding the same estimates, reduces the tediousness associated with the analytical covariance matrix method.
Next, the fitting parameters for the Debye model, which include static permittivity, relaxation time, and specific conductivity, have been extracted as potential indicators of ion concentration and type. A method of judiciously exploiting these indicators, by means of a 3D trajectory plot, is proposed to uniquely identify an ion and infer its concentration from the benchmark data provided in this work.
In addition, for separate ionic aqueous solutions of NaCl, NaNO3, and Na2SO4, concentration- and temperature-dependent parametric models of static permittivity, relaxation time, and specific conductivity that account for physical chemistry and molecular dynamics present in these systems have been developed. These models provide an accurate representation of the radio-frequency (200 MHz to 1 GHz) and microwave (1 to 20 GHz) dielectric spectrum for any of these agriculturally-relevant solutions at a particular concentration and temperature within the ranges studied. The research presented in this dissertation lays a foundation upon which an efficient, real-time, field-deployable, and economically feasible electrical sensing system can be designed for the efficient monitoring and analysis of agricultural run-off.