Previous page     Contents     Next page

Due to the dynamic conditions that prevail in the Earth's interior, minerals experience various chemical, thermal and pressure environments during their histories. Because diffusion rates are low and activation energies are high in silicate minerals, compositional zoning is common and can be used to evaluate the thermal histories of the rocks. Experimentally-measured diffusion coefficients are the necessary input for quantitative modelling. Due to experimental constrains most diffusion experiments are carried out at temperatures higher than 1000 °C. To apply these data to most geological processes they must be extrapolated down to lower temperatures, which is only valid if the diffusion mechanism remains constant over the required temperature range. This makes it necessary to develop a technique which is applicable for the study of slow diffusing species in solids and which can be used to obtain low temperature diffusion data. In this project analytical TEM (ATEM) is used to measure small scale (few µm to sub-µm) diffusion profiles, produced e.g. by low-temperature experiments. Analyses are performed using a Philips CM20-FEG and a very small electron probe (down to ~ 1nm), resulting in ~ 10 nm spatial resolution for quantitative analysis, and concentration profiles down to the nanometer scale. Although other methods are suitable for the measurement of sub-µm diffusion profiles (ion microprobe, Rutherford back scatter spectroscopy (RBS)), the ATEM technique has the advantage of allowing structural characterization with extremely high spatial resolution. Therefore ATEM can be used to study the relation between interface structure (dislocations, inclusions, coherence) and diffusion at the nm-scale, which is not possible with other methods.

The ATEM method is being developed and refined through comparison with electron microprobe (EMPA) data. Diffusion couples were made from Fo-Fo₈₂ single crystals, cut and polished perpendicular to the c-axis, and annealed under controlled temperature and fO₂. The diffusion profiles are measured using ATEM along a line of constant sample thickness such that no absorption correction is required. ATEM profiles fit well with EMPA profiles and are within the precision of the EMPA data (Fig. 3.10-9). The position of the Matano interface, which represents the theoretical interface of diffusion, was calculated after Boltzman and Matano. Diffusion coefficients were calculated using methods dependent and independent of the Matano-interface position. The diffusion coefficients calculated from ATEM data are consistent with those calculated using EMPA profiles. The structural interface, defined by lattice misorientation, is distinct from the Matano Interface, which has not been previously reported. The interface structure is heterogeneous and complex. Samples with only the c-axes oriented show high dislocation densities near the interface on the forsterite side, whereas samples with semicoherent interfaces have few defects.
 

Fig. 3.10-9: Comparison of Mg-Fe interdiffusion profiles in an olivine diffusion couple measured with ATEM and EMPA