Diffusion in glasses studied with X-Ray Photon Correlation Spectroscopy

Diffusion in glasses studied with X-Ray Photon Correlation Spectroscopy

FWF Project NumberP28232-N36
Project Leaderao. Univ. Prof. Dr. Bogdan Sepiol


The movement of single atoms is a fundamental issue in materials science. The fabrication, specific properties and the stability of materials can be significantly improved with knowledge about atomic movement. Numerous properties of materials can be attributed to the single atom motion. Studies of diffusion mechanisms on the atomic scale are challenging despite a number of well-established methods which can be applied to experimentally investigate diffusion in solid state systems. Drawing conclusions from macroscopic measurements on the actual microscopic dynamics is a highly indirect procedure. Due to the relatively long timescales where atomic motion takes place, scattering techniques that rely on a high energy-resolution can only resolve fast processes. The greatest challenge is, however, that only a small number of selected isotopes are accessible to these methods which strongly limits their use. A new method not restricted to certain isotopes and capable of detecting slow diffusion is therefore required. As we have shown in our previous projects these requirements are met by atomic-scale X-ray photon correlation spectroscopy (aXPCS). Former grants served the implementation of XPCS as the preferred method for studying atomic dynamics mainly in crystalline phases. The current project is devoted to studying another very intriguing form of matter, namely the family of glasses. 


Glasses are an active and promising field of research. Understanding disordered solids on a fundamental level and particularly understanding ion conduction in these materials still presents a fundamental challenge. The random formation of structural networks is an important model for explaining the glass forming ability of many materials. The dynamics of basic network components play a key role in decoding puzzling properties of this form of matter like ionic dynamics. Considerable progress with solid-oxide fuel cells, batteries and supercapacitors, in electrochemical sensors and functional polymers has been achieved recently. Even so, the basic concept of ion transport in disordered materials remains poorly understood. This is due to the fact that there is no simple, widely accepted model of transport. Considering the current strong interest in the field and the plethora of experimental and theoretical works it is striking that, in contrast to transport properties in crystalline matter, there is no general consensus on several fundamental questions. It is our aim to shed light on the motion of ions on the atomic level in ionic conducting boron and silica glass and we are confident that the insights gained can be transferred to other ionic glasses.

In our project we will profit from the continuously increasing brilliance of existing synchrotron sources like the upgrade of the ESRF in Grenoble, which was just completed and from new sources like PETRA III in Hamburg.


Beam-induced dynamics in oxide glasses

The correlation rate Γ that is proportional to the diffusion is linear dependent on the dose rate S. Due to the increasing concentration of ions within an alkali borate system the beam-induced effect mitigates.

Previous studies have shown that various materials react differently to the flux of the x-ray beam [1][2]. In contrast to metallic bonding showing no dependencies, oxide glasses exhibit a linear increase of the beam-induced dynamics with regard to the dose rate (flux per second and atom).

The investigation of diverse alkali borate glasses with different alkali concentrations could reveal that the beam-induced effect is decreasing with increasing alkali content, which goes along with the measured glass transition temperatures Tg reflecting the rigidity of the atomic bonds. With an increasing Tg, the beam-induced dynamics of the alkali borate system with increasing alkali content decreases.

In addition to that we could deliver a diffusion model based on beam-induced dynamics measurements over a broad spectrum of the wave vector transfer q. We suggest interactive Brownian motion: Whilst taking into account the effect of De-Gennes, the atoms overcome a distance ~1/q within a certain time τ by infinitesimal small steps causing a stationary atomic diffusion inside the illuminated volume of the sample.

It should be mentioned that the beam-induced effect has negligible low impact on the structure of the material. Thus, the effect could be used for indirectly measuring material-specific properties, e.g. bonding properties. 

The results of this work have been published in: K. Holzweber, C. Tietz, T.M. Fritz, B. Sepiol and M. Leitner: Beam-induced atomic motion in alkali borate glasses. Phys. Rev. B 100 (2019), 214305, DOI: 10.1103/PhysRevB.100.214305.  

[1] B. Ruta, F. Zontone, Y. Chushkin, G. Baldi, G. Pintori, G. Monaco, B. Ruffle, and W. Kob, Sci. Rep. 7, 3962 (2017). 

[2] M. Leitner, M. Stana, M. Ross, and B. Sepiol, “Acceleration of atomic dynamics due to localized energy depositions under Xray irradiation,” (2015), arXiv:1510.01918 [cond-mat.mtrl-sci].

Diffusive dynamics in an amorphous superionic conductor

Charge diffusion coefficients determined in this work (empty blue points) and from [1] (filled green points) with extrapolation (green line), aXPCS diffusion coefficients obtained for full x-ray flux (dark red squares) and with flux attenuated to 25 \% (light red squares), as well as viscosity diffusion coefficient fitted to values measured around 700K in [2] (brown line). The dashed vertical line marks the glass transition temperature.

The general feasibility of aXPCS to study ternary, amorphous substances have proven in works within the scope of a previous project (P22402-N20). Building on this works we extended the study on alkali-borate glasses as a prototypical case of amorphous fast-ionic conductors. The presence of beam-induced dynamics added a layer of complexity onto our aXPCS studies. As an answer to those challenges we introduced a combined approach of impedance spectroscopy (IS) measurements and aXPCS working on different length scales – aXPCS on the microscopic or atomic scale and IS on the macroscopic scale. Since it was established previously that the electric conduction in alkali borate glasses is indeed of ionic nature, it combining aXPCS with IS meant additional access to atomic diffusion.

Temperature-dependent diffusion coefficients from aXPCS and conductivity measurements as well as literature data on the viscosity revealed the presence of different diffusion processes. It was shown that the aXPCS probes the rearrangement of the boron-oxygen network matrix corresponding to the viscosity determining diffusion process, but obtains smaller correlation times due to the acceleration of the beam-induced effect. On the other hand the fast dynamics of the ionic conductivity is found to take place on an arrangement of largely populated and distinct cationic via a vacancy mechanism known from crystalline materials. Diffusion over this alkali sites is largely decoupled from the network matrix and concomitantly the alkali sites remain stable on the timescale of ionic diffusion.   

A dedicated master thesis in our group on the topic of IS experiments provided the know-how and experimental data to complement the aXPCS data collected on synchrotron facilities.

The results of this work have been published in: C. Tietz, T. M. Fritz, K. Holzweber, M. Legenstein, B. Sepiol, and M. Leitner: Diffusive dynamics in an amorphous superionic conductor. Phys. Rev. Research 2 (2020), 043141, DOI: 10.1103/PhysRevResearch.2.043141. The experimental data were published open-source in an online repository for science data: for aXPCS data see DOI: 10.6084/m9.figshare.12651458.v1 and for IS data see DOI: 10.6084/m9.figshare.12651560. The full data of the IS measurements are published in the master‘s thesis of  T. M. Fritz.

[1] F. Berkemeier, S. Voss, Á.W. Imre, and H. Mehrer, J. Non-Cryst. Solids 351, 3816 (2005)

[2] S. V. Stolyar, L. V. Grishchenko, and G. A. Sycheva, Glass Phys. Chem. 32, 293 (2006)