Table of Contents:
In the scope of this doctoral thesis, continuing previous studies the Plasma Charge Attachment Induced Transport (CAIT) experiment was devoloped and applied to study ion conducting glasses. Additionally, Time-of-Flight Secondary Ion Mass Spectrometry and theoretical calculations on the basis of the Nernst-Planck-equation and the Poisson equation were employed. The determination of ionic conductivities, concentration dependent diffusion coefficients as well as the potential energy landscape of a single glass in form of a populated site energy distribution (PSED) was archived.
The principle of CAIT is based on the controlled attachment of charge carriers to the surface of a solid electrolyte. Previously, versions employing either alkali ion or electron beams have been developed. Each one is generated by thermionic emission. Thus, high vacuum conditions are required. There are two measurement modi. Conductivity studies are performed by recording current-voltage curves at several temperatures. Constant Voltage Attachments (CVA) aim to generate modified concentration depth profiles, which can be demonstrated ex situ by ToF-SIMS. Therefore, the attachment is performed under steady experimental conditions, usually on longer time scale. Instead of a thermionic emitter and according to the name, the Plasma CAIT experiment employs a plasma as ion source. It is ignited by focussing high intense fs laser pulses. This results in a dielectric breakdown of the residuing gas around the focal point. A partial separation of charge carriers is accomplished by applying a static electric field to the plasma. This way, charge carriers of the desired polarity can be attached to the surface of a sample of interest. Previous to this thesis, there was already access to the activation energies for ionic transport. However, absolute conductivities deviated more than an order of magnitude from literature values. Thus, a calibration against a reference was required. Further on, there was only access to plasma ignition at ambient air.
The first project of this thesis was a continuation of the method development previously started. A vacuum-tight, evacuable housing around the measurement setup was constructed. Filling this chamber with any desired gas enables to control the chemical identity of the attached species. From the beginning, the irradiated area of a sample has been defined by a mask. Yet previously electrically grounded, now it was set to a floating potential. A model system, mimicing an ionconducting sample, was employed for the charaterization of this setup. It consisted of a pure metal electrode and resistors in situ exchangeable and of known values, connected in series. The resulting current-voltage curves could be explained by a serial connection of an apparatus resistance (the plasma) and the respective resistor. Measurements at several pressures showed that a reduced pressure enables the extraction of a larger amount of charge carriers from the plasma. Afterwards, the vast mayority of experiments were performed at hydrogen pressure of 200 mbar, in order to study proton attachment. Hence, the calibration-free measurement of ionic conductivities was enabled.
The plasma resistance is treated as an adjustable parameter in order to obtain the best linear regression in the Arrhenius fit. The proof of principle was shown on a previously wellcharacterized D263T glass. To conclude this project, the first CVA experiments by the means
of Plasma CAIT were also performed on D263T. A resulting replacement profile agreed qualitatively with a previous profile generated by cesium ion attachment. Yet, it was concluded that the detected 1H+ signal in SIMS spectra partially results from adsorbed residual gas in the vacuum chamber. Thus, a further CVA experiment on D263T and all the following CVA experiments have been performed in deuterium atmosphere instead. The second project of the thesis dealt with lithium aluminum germanium phosphate glasses. An activation energy of 0.73 eV was determined for long range lithium ion motion. A CVA experiment was performed, analyzed quantitatively and simulated by the means of NPP. This resulted in the determination of a populated site energy distribution with a full width half maximum of 113 meV. This value is significantly lower than the ones of two sodium ion conducting glasses studied previously to this thesis. Hence, there is a first experimental proof of PSEDs in glasses being a material property.
The third project continously dealt with alkali aluminum germanium phosphate glasses. In the glass preparation, lithium was systematically substituted by all the other alkali cations. A conductivity study showed a systematic decrease of conductivities and an increase of activation energies in the order from litium to cesium. In addition to Plasma CAIT, conductivity measurements were also performed in a two-electrode DC setup as well in hydrogen atmosphere as under medium vacuum conditions. For this, sputter coated platinum electrodes were employed. In the respective limit of certainty, the results of all three measurements coincided well. Any conceivable blocking, current reducing behavior, e.g. the formation of double layers does not disrupt two electrode measurements. Conductivity measurements are enabled as long as the modification of the materials studied remains neglible. Changes are noticed by non-linear behavior in current-voltage curves and/or the resulting Arrhenius plot.
Additionally, CVA-experiments attaching deuterons were performed on the MAGP glasses. Qualitatively, the shape of deuterium concentration profiles determined got more shallow in the order LAGP, NAGP, KAGP, RAGP. This trend is explained by an increasing deuterium diffusion coeffient relative to the respective alkali cation.