TWIN DOMAIN SWITCHING IN NEIGHBORITE (NAMGF3) DURING OSCILLATING DEFORMATION: POTENTIAL IMPLICATIONS FOR INTRINSIC SEISMIC ATTENUATION OF THE LOWER MANTLE

With an orthorhombic perovskite structure, bridgmanite is prone to twinning under deviatoric stress [1, 2]. Twinning introduces internal friction, thereby affecting seismic attenuation [3]. A recent seismic study shows that under eastern Eurasia, lower mantle quality factors (Q) for P- and S-waves are on the order of 3000 and 1300, respectively, much higher than the global average given by PREM (~800 and 300, respectively) [4]. Attenuation (1/Q) may be divided into two broad categories: 1) intrinsic attenuation, which originates from the crystal structure under a given P, T condition at seismic frequency and 2) extrinsic attenuation due to small-scale scattering from the so-called marble cake texture. While low intrinsic attenuation implies low temperature, low extrinsic attenuation implies lack of texture. It is essential to distinguish the two sources to gain physical insights into the dynamic processes. However, since bridgmanite is stable under pressures well above 22 GPa, detailed experimental investigation on this mineral is currently very difficult. Here we report a preliminary study on intrinsic attenuation of an isostructural analogue to bridgmanite – neighborite (NaMgF₃) – at ID06 of ESRF. We used monochromatic radiation of 53 keV with a rotating, long aspect-ratio 2D Pilatus CdTe detector, at the sample-detector distance of ~4 m. With this unique setup, the characteristic triplet (200, 112, and 020) of NaMgF₃ was completely resolved without any overlap. Radiographic imaging and ultrasonic techniques were used simultaneously to resolve small strain. As the first feasibility study, we show that intensities of the 200 and 020 lines exhibit reversable switching due to flipping twin domains at an oscillating strain amplitude of ~ 10^-4 s^-1, with oscillation periods of 1200 – 3600 s. These preliminary results demonstrate the feasibility of the experimental approach. With future efforts to improve imaging and ultrasonic data collection, we expect to be able to resolve strain to below 10^-5 s^-1. With focused monochromatic beam, we expect to collect diffraction data at periods significantly shorter than 1000 s.

References

[1] E.K.H. Salje (1990) Phase transitions in ferroelastic and co-elastic crystals.

[2] Y. Wang, et al. (1990) Science, 248, 468.

[3] Carpenter, M.A., Z. Zhang (2011) Geophys. J. Int., 186, 279.

[4] B.L. Zhang, et al. (2019) Earth Planet. Phys., 3, 537.