Inelastische Neutronenstreuung bei hohen Temperaturen und Drücken zur Aufklärung der Lösungs- und Transportmechanismen von Wasser in wasserhaltigen Silikatschmelzen
Final Report Abstract
The aims of the current project are to construct a high temperature high pressure sample environment for neutron time-of-flight experiments and to study water dynamics in hydrous silicate melts under high temperature and high pressure conditions with quasielastic neutron scattering techniques. The addition of water to silicates causes a drastic change of the melt properties. The understanding of the relaxation and transport mechanisms in silicate melts is very important for many geological processes, especially active volcanism. Water is known to partially react with silicates upon its dissolution, resulting in two different species in the melt: OH-groups and molecular water. The knowledge of the water dynamics represents an essential key to understand the melt properties. However, its dissolution and transport mechanisms are still not fully understood. Neutron scattering techniques give access to investigate dynamics on microscopic time and length scales. The intrinsic q resolution of the quasielastic neutron scattering allows diffusion mechanisms to be studied in great detail. Neutron scattering also provides the possibility to perform a contrast variation via isotope substitution. In the case of water, a substitution of H2O by D2O enables to extract pure proton signals. To study the hydrous silicates at temperatures higher than their glass transition temperatures, high pressure in the order of 1-2 kbars is simultaneously required to suppress water evaporation. Therefore, a high temperature high pressure sample environment was built which is optimized for the neutron time-of-flight spectrometer TOFTOF at FRM II. Nb1Zr alloy is chosen as the cell material since it provides a sufficient mechanical strength at elevated temperatures and has also an extremely small incoherent neutron scattering cross section. Hence, an acceptable signal-to-background ratio of about 10:1 is achieved within the elastic and quasielastic region, even with 24 mm cell material in the beam. An internal heating setup is used to heat up the samples. With such setup the sample environment provides a temperature range from ambient temperature up to 1500 K at pressures up to 2 kbar at the sample position with a sample volume around 1 cm3. The realization of the high temperature high pressure sample environment opens a new possibility of direct observation of dynamics in hydrous silicate melts using quasielastic neutron scattering techniques. Hydrous Na2Si3O7, NaAlSi3O8, and SiO2 melt have been studied. In the hydrous NaAlSi3O8 and SiO2 system, the proton dynamics is surprisingly slow. No resolvable quasielastic broadening or decay of the intermediate scattering function has been observed with the instrumental energy resolutions available on TOFTOF at the highest measured temperature. The lower boundary value of the diffusion coefficients is on the order of 10^-10^ m2s-1 at these temperatures. An unusual relaxation behaviour of the proton in hydrous sodiuin trisilicate melt with a fast and a slow component has been observed. In the S(q,t) the fast component exhibits extreme stretching which might be described by a logarithmic like decay to a plateau value around 20-25 ps. The further decay of this constant value towards zero is out of the accessible time window at TOFTOF. Through a careful analysis of the q and temperature dependence of the plateau values, an attribution of the fast/slow relaxations to different proton environments cannot be satisfied. Within the neutron experiment results no evidence of different dynamics due to different water species has been identified. The logarithmic like decay is a signature of high orders of glass transition, which can occur in systems having different competitive arrest mechanisms, as predicted by mode coupling theory. This has been observed in binary hard spheres mixtures with a sufficient large size disparitiy as well as sodiuin trisilicate melt under certain conditions. Tins could be also the case in hydrous sodium silicate melts observed here. To verify such interpretation it is necessary to further study the final decay of the intermediate plateau value f2(q) to zero. This was only achieved recently by measurements on a unique neutron near backscattering spectrometer BASIS at SNS in combination with the pressure cell. A detailed analysis of the transport mechanism is in progress.
