Minimizing and quantifying mis-indexing in electron backscatter diffraction (EBSD) determinations of antigorite crystal directions

- EBSD-derived antigorite CPO depends on sample preparation and MAD values.
- XY-samples give the most reliable antigorite crystallographic orientations.
- Mis-indexing issues of antigorite EBSD measurements can be minimized by using MAD values <0.7°.
- Mis-indexing of a- and b-axes occurs even for low MAD values.

Antigorite is common in hydrated mantle domains and commonly shows strong alignment either due to deformation or growth in a preferred orientation. The alignment of antigorite imparts a strong anisotropy to the host rock affecting physical properties. A quantitative analysis of how antigorite affects these properties requires a reliable measurement of the crystallographic preferred orientation of antigorite, and EBSD analysis is the most widely used technique. Potential problems include (i) mis-indexing, which can add significant uncertainties to the results; and (ii) sample preparation, which may affect the measured CPO in particular for automated mapping. Combining results derived from X-ray goniometry, EBSD and U-stage techniques with FIB-TEM analysis leads to the following conclusions concerning these two issues. (i) There is a significant issue with mis-indexing a- and b-axes due to rotational similarities about the c-axis. Similar problems may also affect the c-axes measurements but this is less significant than the a- and b-axes when data are filtered using lower MAD values. Filtering using MAD values of <0.7° can significantly change the resulting CPO. (ii) Sample preparation can also affect the measured CPO: sections prepared parallel to the foliation show only minor differences with MAD values of <2.0°. Mis-indexing problems can be minimized by using an MAD value of <0.7° and analysing thin sections cut parallel to the foliation.