Pressure-induced slip-system transition in forsterite: Single-crystal rheological properties at mantle pressure and temperature

Abstract

Deformation experiments were carried out in a Deformation-DIA high-pressure apparatus (D-DIA) on oriented Mg 2 SiO 4 olivine (Fo100) single crystals, at pressure (P) ranging from 2.1 to 7.5 GPa, in the temperature (T) range 1373-1677 K, and in dry conditions. These experiments were designed to investigate the effect of pressure on olivine dislocation slip-system activities, responsible for the lattice-preferred orientations observed in the upper mantle. Two compression directions were tested, promoting either [100] slip alone or [001] slip alone in (010) crystallographic plane. Constant applied stress (σ) and specimen strain rates (ε . ) were monitored in situ using time-resolved X-ray synchrotron diffraction and radiography, respectively. Transmission electron microscopy (TEM) investigation of the run products reveals that dislocation creep assisted by dislocation climb and cross slip was responsible for sample deformation. A slip transition with increasing pressure, from a dominant [100]-slip to a dominant [001]-slip, is documented. Extrapolation of the obtained rheological laws to upper-mantle P, T, and σ conditions, suggests that [001]-slip activity becomes comparable to [100]-slip activity in the deep upper mantle, while [001] slip is mostly dominant in subduction zones.These results provide alternative explanations for the seismic anisotropy attenuation observed in the upper mantle, and for the "puzzling" seismic-anisotropy anomalies commonly observed in subduction zones.

Experimental deformation of olivine single crystals at mantle pressures and temperatures

“…This geometry, previously used by Raterron et al (2007), ensures that both a-sample and c-sample (i.e., a and c dislocations) experience the same stress field during deformation. It allows direct comparison of sample deformation at run conditions, i.e., of a and c dislocation relative activities, with no need of knowing precisely the applied stress.…”

“…These techniques, previously used successfully (e.g., Li et al (2006a), and (2006b); Raterron et al (2007)), allow in situ measurement of the pressure and stress applied on specimens, and resulting specimen strain rates. Pressure and applied stress were measured from the quantification of x-ray peak shifts in diffraction patterns arising from stress calibrants: polycrystalline olivine and/or alumina, and/or MgAlB B 2 B B OB B 4 B B -spinel pellets placed in the compression column and separated from the sample(s) by Ni foils, and/or polycrystalline alumina pistons at sample ends ( Figure 1).…”

“…Stress uncertainty using these techniques is fairly large, since it depends on the accuracy on the measured d spacing -i.e., on the position of x-ray diffraction peak maxima and shifting toward higher energy on compression (smaller d spacing) -which in turn depends on several experimental factors (see Raterron et al (2007) for details).…”

“…These images were recorded after magnification on a CCD camera (for details, see Vaughan et al (2000)). Sample and calibrant end positions appear as black lines on the radiographs (e.g., Raterron et al (2007)). These lines result from x-ray absorption by 25-µm thick metal-foils (strain markers: here Ni foils).…”

“…Olivine slip system relative activities have thus been extensively studied, although mostly at relatively low hydrostatic pressure (P < 3 GPa) because of experimental limitations (Raleigh, 1968;Carter and Avé Lallemant, 1970;Phakey et al 1972;Kohlstedt and Goetze, 1974;Poirier, 1975;Kohlstedt et al, 1976Kohlstedt et al, , 1980Durham et al, , 1979Gueguen, 1979;Jaoul et al, 1979;Darot and Gueguen, 1981;Gueguen and Darot, 1982;Mackwell and Kohlstedt, 1985;Ricoult and Kohlstedt, 1985;Bai et al, 1991;Kohlstedt, 1992a, 1992b;Jin et al, 1994). At P < 3 GPa, mantle temperature (T > 1173 K), moderate differential stress and in dry condition, olivine [100] (010) Yet, recent deformation experiments carried out at high pressure on forsterite (Fo100) crystals (Couvy et al, 2004;Raterron et al, 2007), as well as theoretical studies based on first-principle calculations (Durinck et al, 2005;Durinck et al, 2007), show that olivine c-slip may dominate deformation in the P and T ranges of the deep upper A c c e p t e d M a n u s c r i p t 3 mantle. This may promote a shear-parallel slow-velocity [001] axis, which could explain the seismic-velocity attenuation with depth (Mainprice et al, 2005) observed in the upper mantle.…”