Olivine low-angle grain boundaries (LAGBs) influence the upper mantle properties, but the role of geometry in their structure and chemical composition remains largely unexplored. Here, we characterise and compare three tilt LAGBs with a 4.5° misorientation angle but different misorientation axes in a mantle xenolith. Electron backscatter diffraction (EBSD) was used to characterise three olivine grains, containing one LAGB aligning with the (001)100 slip system, and two LAGBs with (010)100 slip. Within the (010)100 LAGBs, transmission electron microscopy (TEM) shows dislocation cores regularly spaced at 5.8 nm. Direct dislocations imaging in the (001)100 LAGB was hindered by lamella orientation. Atom probe tomography (APT) reveals segregation of Al, Ca, Fe, H and Ti to the LAGBs, accompanied by Mg depletion. In the (010)100 LAGBs, the segregated elements are concentrated in linear arrays (~ 5.8 nm spacing), consistent with segregation to dislocations. In the (001)100 LAGBs, although segregated elements appear evenly distributed, 2D profiles show regularly spaced features at ~ 4.8 nm along the boundary, indicative of dislocation spacing. Interfacial excess calculations reveal differences in elemental segregation between boundaries. The (001)100 LAGB has greater H segregation, while the LAGB in the seemingly smaller grain exhibits decreased Al, Ca and Fe enrichment. These findings suggest that LAGB geometry influences elemental segregation. Because these geometries are associated with specific mantle fabrics, the segregation patterns may influence differences in phase transitions, creep behaviour, electrical conductivity, and seismic properties in the upper mantle.
Building similarity graph...
Analyzing shared references across papers
Loading...
Dimitris Dimitriou
University of Zurich
Steven M. Reddy
Planetary Science Institute
Denis Fougerouse
Planetary Science Institute
Contributions to Mineralogy and Petrology
University of Oxford
Imperial College London
Commonwealth Scientific and Industrial Research Organisation
Building similarity graph...
Analyzing shared references across papers
Loading...
Dimitriou et al. (Wed,) studied this question.
synapsesocial.com/papers/69db38534fe01fead37c6862 — DOI: https://doi.org/10.1007/s00410-026-02318-9