Dislocation migration in terrestrial and impact deformed olivine

Introduction: Shock deformation arises from the hyper-velocity impact from two or more bodies. It is a key mechanism for planetary formation in the primordial solar system. Olivine is the fundamental rock-forming mineral and the major building-block for a rocky planetary body. In a differentiated planet, Mg-rich olivine is crystallized primarily from the magma forming planet’s mantle. Its orthorhombic crystal structure provides well-defined ways (slip systems) to accommodate the deformation stress and its high melting point makes it a perfect candidate to record the deformations for extreme condition, e.g., shock metamorphism. The interplanetary collision unloads the pressure instantaneously, resulting in an extreme strain rate deformation (up to10⁶/s), with the pressure decaying from the epic centre. Thus, in the same body, it is possible to have both high-shocked (>45GPa) and low-shocked rocks (<20GPa)^1,2. The dislocations induced by shock are distributed randomly in the crystal with no preferred orientation, forming shock mosaicity^1,2,3. It is a crucial petrographic texture to distinguish shock from other plastic deformation (e.g., tectonism). Achondrite meteorites, such as those from the Moon, Mars, or ureilites (Figure 1), originate from fully or partially differentiated planetary bodies. These meteorites record both shock deformation from ejection events and information about their parent bodies. 

Methods and samples: In this work, we used the synchrotron-based dark-field X-ray microscopy (DFXM) coupled with other imaging method, electron-backscattered diffraction (EBSD) and 2 dimensional X-ray diffraction (2D-XRD),  to systematically explore the microstructure development in differently deformed olivine, including non-shocked terrestrial olivine peridotite and kimberlite, to low-shock ureilite EET 96042, and to highly shocked Martian shergottite NWA 7721 and Martian chassignite NWA 2737. By using a line-focus beam, we scanned grain in different layers in sample z-dimension with DFXM, offering maximum resolution of 35 nm, 105 nm, and 500 nm in sample x, y, and z dimension. In the end, it allowed the in-situ reconstruction of 3D deformation volume of the examine grain non-destructively.

Results: DFXM shows superior power of resolving microstructures formed by the low-angle boundaries, revealing a distinctive difference between non-shock and shock olivine. We report the development of the “dislocation networks” converging to or diverging from the point of failure of the fractures, forming the incipient shock mosaicity, corresponding well with previous work of Li et al. (2021, 2023, and 2025)^3,4,5. In detail, the dislocation distributions in the 3D grain volume that 1) “dislocation network” formed by very-low-angle misorientation boundaries (<0.1 º) and 2) incipient subdomain walls formed by low-angle misorientation boundaries (> 0.3º). These features are not observed in the non-shocked samples. Furthermore, we performed autocorrelation using 2D fast-Fourier transformation on the misorienation maps. We discovered the disrupted and broadened slip-band features at spacing of 10 µm in each layer. By comparing the observation with terrestrial deformed olivine, we conclude they are remnant features from the crustal process rather than impact. Our research highlights the potential of using microstructures to understand different deformation features using observations from multi-techniques at multi-scales. It sheds light into deconvoluting the shock features from parent body process, offering a novel way to decode meteoritic materials and their parent body deformation history.

 

Reference:  

[1] Fritz, J., Greshake, A., and Fernandes, V.A. (2017). MAPS, 52, 1216–1232.  [2] Stöffler, D., Hamann, C., and Metzler, K. (2018) MAPS, 53, 5–49. [3] Li, Yaozhu, McCausland, P.J.A., and Flemming, R.L. (2021) MAPS, 56, 1422-1439. doi:10.1111/maps.13706. [4] Li Y., McCausland P. J. A., Flemming R. L., and Hetherington C. J. 2023. AmMin. 108:1897–1905. [5] Li, Y., McCausland, P. J. A., Flemming, R. L., & Osinski, G. R. (2025 MAPS, 60(2), 347-370.