PosterPDF Available

Diffusion of H-bearing species in silicate glasses at low temperatures: development of a new experimental technique

Authors:
InputInitial speciationFinite difference schemeSpeciationResultSchematic diffusion modeling routine with implemented speciation assuming instantaneous speciation reaction. Alternative routines are also tested. Numerical modellingDiffusion modelling is performed by applying a numerical solution (finite difference). This allows us to set up initial and boundary conditions that describe the real scenario as well as possible. Also, we can implement and test different diffusion [4]models (e.g. Zhang et al., 1991). Concentration profiles that were measured prior to diffusion anneals are used as “starting“ profile of time t = 0. In models with speciation the profiles (H at. %) are transformed to both H Om and OH on basis of 2[5]known models. The partitioning can be varied by changing the hypothetical equilibrium speciation constant Q. Distinct diffusion coefficients can be assigned to each species, different concentration dependencies and alternative speciation be-haviour are tested.One set of experiments was performed at 1 atmosphere pressure and 350 °C. The hydrogen concentration in the thin layer, that constituted a plateau in the beginning, decreased evenly while water diffused into the dry glass substrate. The profile shape in the thin film reveals two aspects: (i) very little H is lost from the surface of the sample, (ii) transport in the film is relatively fast (flat gradients). The low solubility of H in glass at 1 atmosphere controls the concentration of H at the film substrate interface; the concentration profile forms a tail into the substrate from this low concentration at the interface. Thus, the profile from the 3 day experiment was taken as the starting point for modelling the profiles from the longer experiments. Several speciation models were tested. The best fit for this diffusion profile was archieved using the diffusion model by Zhang et al. (1991). Simpler models that do not include a speciation reaction are unsuitable to describe the entire profile shape. However, this does not need to be true for experiments at 2 kbar.One experimental set was performed at 2kbar pressure and 350 °C. The diffusion anneal was carried out in an Internally Heated Pressure Vessel (IHPV). Black squares represent the initial H O concentration profile, that was 2measured before the experiment. The grey dots are the diffusion profile after the experiment. The concentration plateau of the thin layer decreased evenly, while water diffused into the substrate. However, the plateau spreads into the dry glass. The slope of the diffusion front consists of a steep section followed by a less inclined tail. Another set was performed at 2 kbar pressure and 350 °C using cold seal pressure vessels. After 4 days a pro-nounced slope into the dry glass substrate had developed. However, the water concentration in the thin layer changed only negligibly, which results in a violation of mass balance. More elaborate capsule designs to buffer the diffusive transport of hydrogen through the capsule material or the use of isotopic tracers may solve this problem.[1] Dohmen, R. et al. (2002), European Journal of Mineralogy 14, 6, S. 1155–168.[2] Watson, E. B.; Dohmen, R. (2010), Reviews in Mineralogy and Geochemistry 72, 1, S. 61–05.[3] Becker, H.-W.; Rogalla, D. (2016), H. Fritzsche, J. Huot und D. Fruchart (Hg.) S. 315–36.[4] Zhang, Y.; Stolper, E. M.; Wasserburg, G. J. (1991),Earth and Planetary Science Letters 103, 1-4, S. 228–40.[5] Stolper, E. (1982): The speciation of water in silicate melts. In: Geochimica et Cosmochimica Acta 46, 12, S. 2609–620.
Ÿ Immobilization of nuclear waste in glassŸ Seafloor weatheringŸ Durability of thin film (solar cells and electronic devices)Ÿ Age dating of volcanic glass artefactsŸ Explosivity of volcanic glass shards and tephraGlass hydrationHydrous glass is produced by annealing glass powder with DI-water in AuPd-capsules at 1200 °C and 2 kbar for 24 hrs in an Internally Heated Pressure Vessel (IHPV). The hydration is terminated by rapid quench. The hydrated glass is used as target material for the diffusion couple production. Silicate glasses of geological and commercial interests are used in this study. Some of these are synthesised by mixing oxides, hydroxides and carbonates and melting at 1500 °C. Grinding and repeated melting guarantees compositional homogeneity. Melts are either quenched in air or water. Glass synthesis[1],[2]Pulsed laser deposition (PLD)Diffusion couples are produced by applying hydrated silicate glass thin films onto a dry glass substrate by PLD. The pulsed laser beam (l = 193 nm) of an Excimer laser hits the rotating target (hydrated glass) and produces a plasma in an evacuated chamber. The plasma is deposited as thin layer (200 nm) onto the dry glass. This method enables us to perform diffusion experiments without the presence of free water at the interface, that would lead to alternative processes, e.g. dissolution and precipitation. Therefore, concentration profiles may be directly attributed to diffusion.153+ NRRA is used for measurements of hydrogen depth profiles. This analytical technique uses Nions 111512that initiate a nuclear reaction by hitting H at energies of 6.385 MeV. The reaction H(N,ag)C is highly energy sensitive with a narrow energy window of 1.8 keV, which provides a depth resolution of few nm. This feature is essential for nanometre-scale profiles that result from low temperature experiments. Detection of g-particles (4.4 MeV) and correlation of these with the integrated ion beam current gives a direct measure of the hydrogen content. Depth profiles are produced by increasing the ion beam energy stepwise. The method is independent of chemical bondings; total hydrogen is detected. No standards are needed and the non-destructive nature of the technique enables us to obtain initial concentration profiles prior to experiments and to observe the development of diffusion profiles in one sample.[3]Nuclear Resonant Reaction Analysis (NRRA)Diffusion of hydrogen bearing species in silicate glasses plays a considerable role in various applications that cover a broad range of scientific as well as commercial domains. However, diffusion of these species at low temperatures have been studied only to a limited extent because of experimental difficulties. Here we investigate the diffusion of hydrogen bearing species in silicate glasses below the glass transition temperature using the following novel tools: (I) Production of diffusion couples by applying hydrated silicate glass thin films onto a dry glass substrate by pulsed laser deposition (PLD). (II) Measurement of nanoscale concentration profiles that result from diffusion at low temperatures using nuclear resonant reaction analysis (NRRA). (III) Modelling of unconventional profile shapes using numerical methods (finite difference).1,21,2.23Bißbort, T., Chakraborty, S., Becker H.-W, Fanara, S. (1) Institut für Mineralogie, Ruhr-Universität Bochum, Germany (2) RUBION, Ruhr-Universität Bochum, Germany (3) Institut für Mineralogie, Georg-August Universität Göttingen, Germany Diffusion of H-bearing species in silicate glasses at low temperatures - development of a new experimental technique -
Institute of Geology, Mineralogy & Geophysics
RUBION Central Unit for Ionbeams and Radionuclidesbeam energy [MeV]
g-yields [counts/ 0.3 mC]
6.3 6.4 6.5 6.6 6.7 15,00010,0005,0000 g-energy [MeV]
counts/ 0.3 mC
0.0 2.0 4.0 6.0 8.0 10.0 10,0005,0000 ROIrotating target holderlaser beamtargetplasmasubstrate
0 100 200 300 400 500 6002.52.01.51.00.5distance [nm]
H O [wt. %]2
An50Di50 | T = 350 °C | P = 1 atmosphereinitialprofile 200 300 1.00.5
3 days6 days15 days24 days33 days42 days
0 50 100 150 200 250 300 1.81.41.00.60.2distance [nm]
H O [wt. %]2
t = 96 hrs
initialprofile An50Di50 | T = 350 °C | P = 2 kbar | IHPV
no c-dep. H O+OH 2linear c-dep.
0 100 200 300 4003.02.52.01.51.00.50distance [nm]
H O [wt. %]2
initialprofile
t = 96 hrs
An50Di50 | T = 350 °C | P = 2 kbar | hydrothermal
2D(H Om) = 1E-20 m /s22D(OH) = 9E-21 m /s(preliminary)2D(H Om) = 2E-20 m /s22D(OH) = 8E-22 m /s(preliminary)
2 kbar experimental resultsMethodsReferencesApplicationsIntroduction1 atm experimental results
0 2 4 6 8 10 54321total H O [wt. %]2
species concentration [wt. %]
1.00.80.60.40.2
spproportionecies
HOm2HOm2OHOHQ = 0.12H Om t=02OH t=0H Ot t=02
H Ot t=39d2 H Ot fit2H Om fit2OH fit
2D(H Om) = 3.0e-21 m /s22D(OH) = 8.0e-25 m /s(preliminary)
0 100 200 300 400 500 1.41.21.00.80.60.40.2
concentration [wt. %]
distance [nm]0 100 200 300 400 500 1.41.21.00.80.60.40.2
concentration [wt. %]
distance [nm]
2D = 1.15e-21 m /s2D linear = 1.5e-21 m /s(preliminary)
H Ot t=02
H Ot t=39d2
H Ot fit linear 2c-dependenceH Ot fit2no speciation
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.