High-temperature creep in a Ni2GeO4: a contribution to creep systematics in spinel

 High-temperature creep behavior in Ni₂GeO₄ spinel was investigated using synthetic polycrystalline aggregates with average grain sizes ranging from submicron to 7.4 microns. Cylindrical samples were deformed at constant load in a gas-medium apparatus at temperatures ranging from 1223 to 1523 K and stresses ranging from 40 to 320 MPa. Two deformation mechanisms were identified, characterized by the following flow laws: where σ is in MPa, d is in μm and T is in Kelvin. These flow laws suggest that deformation was accommodated by dislocation creep and grain-boundary diffusion (Coble) creep, respectively. A comparison with other spinels shows that an isomechanical group can be defined for spinels although some differences between normal and inverse spinels can be identified. When creep data for olivine and spinel are normalized and extrapolated to Earth-like conditions, spinel (ringwoodite) has a strength similar to olivine in the dislocation creep regime and is considerably stronger than olivine in the diffusion creep regime at coarse grain size. However, when grain-size reduction occurs, spinel can become weaker than olivine due to its high grain-size sensitivity (Coble creep behavior). Analysis of normalized diffusion creep data for olivine and spinel indicate that spinel is weaker than olivine at grain sizes less than 2 μm. Received: 18 June 2000 / Accepted: 3 April 2001     

Synthetic MgGa2O4-Mg2GeO4 spinel solid solution and β-Mg3Ga2GeO8: chemistry, crystal structures, cation ordering, and comparison to Mg2GeO4 olivine

 Structural parameters and cation ordering are determined for four compositions in the synthetic MgGa₂O₄-Mg₂GeO₄ spinel solid solution (0, 8, 15 and 23 mol% Mg₂GeO₄; 1400 °C, 1 bar) and for spinelloid β-Mg₃Ga₂GeO₈ (1350 °C, 1 bar), by Rietveld refinement of room-temperature neutron diffraction data. Sample chemistry is determined by XRF and EPMA. Addition of Mg₂GeO₄ causes the cation distribution of the MgGa₂O₄ component to change from a disordered inverse distribution in end member MgGa₂O₄, ^[4]Ga = x = 0.88(3), through the random distribution, toward a normal cation distribution, x = 0.37(3), at 23 mol% Mg₂GeO₄. An increase in a_o with increasing Mg₂GeO₄ component is correlated with an increase in the amount of Mg on the tetrahedral site, through substitution of 2 Ga³⁺⇄ Mg²⁺+Ge⁴⁺. The spinel exhibits high configurational entropy, reaching 20.2 J mol⁻¹ (four oxygen basis) near the compositional upper limit of the solid solution. This stabilizes the spinel in spite of positive enthalpy of disordering over the solid solution, where ΔH _D  = αx + βx ², α = 22(3), β = −21(3) kJ mol⁻¹. This model for the cation distribution across the join suggests that the empirically determined limit of the spinel solid solution is correlated with the limit of tetrahedral ordering of Mg, after which local charge-balanced substitution is no longer maintained. Spinelloid β-Mg₃Ga₂GeO₈ has cation distribution ^M1[Mg_0.50(2)Ga_0.50(2)] ^M2[Mg_0.96(2)Ga_0.04(2)] ^M3[Mg_0.77(2) Ga_0.23(2)]₂ (Ge_0.5Ga_0.5)₂O₈ (tetrahedral site occupancies are assumed). Octahedral site size is correlated to Mg distribution, where site volume, site distortion, and Mg content follow the relation M1<M3<M2. The disordered cation distribution provides local electrical neutrality in the structure, and stabilization through increased configurational entropy (27.6 J mol⁻¹; eight oxygen basis). Comparison of the crystal structures of Mg_{1+ N} Ga_{2−2 N} Ge _N O₄ spinel, β-Mg₃Ga₂GeO₈, and Mg₂GeO₄ olivine reveals β-Mg₃Ga₂GeO₈ to be a true structural intermediate. Phase transitions across the pseudobinary are necessary to accommodate an increasing divergence of cation size and valence, with addition of Mg₂GeO₄ component. Octahedral volume increases while tetrahedral volume decreases from spinel to β-Mg₃Ga₂GeO₈ to olivine, with addition of Mg and Ge, respectively. Furthermore, M-M distances increase regularly across the join, suggesting that changes in topology reduce cation-cation repulsion. Received: 9 November 1998 / Revised, accepted: 3 August 1999     

Thermochemistry of phases in the system MgGa2O4-Mg2GeO4

The enthalpies of solution in molten 2PbO · B₂O₃ of phases synthesized at one atmosphere in the system MgGa₂O₄-Mg₂GeO₄ have been measured. A spinel solid solution, which is stable at 1400 °C from the MgGa₂O₄ end-member to 27 mole percent Mg₂GeO₄, shows endothermic heats of mixing of up to 10 kJ/mole at the solubility limit. The spinelloid phase, Mg₃Ga₂GeO₈, is energetically less stable than a mixture of terminal spinel solid solutions (0.73 MgGa₂O₄·0.27 Mg₂GeO₄(sp)+Mg₂GeO₄(sp)), by 3.63±3.64 kJ/mole. This indicates that the spinelloid is a high-entropy phase.The volume of the spinel solid solution, MgGa₂O₄-Mg₂GeO₄, shows a positive deviation from Vegard's Law. Modeling of the cation distribution in the solid solution indicates that this V is due to a change in the spinel type from inverse towards normal as the Mg₂GeO₄ content increases.     

A computer simulation study of OH defects in Mg2SiO4 and Mg2GeO4 spinels

Classical atomistic simulation techniques have been used to investigate the energies of hydrogen defects in Mg₂SiO₄ and Mg₂GeO₄ spinels. Ringwoodite (γ-Mg₂SiO₄) is considered to be the most abundant mineral in the lower part of the transition zone and can incorporate large amounts of water in the form of hydroxyls, whereas the germanate spinel (γ-Mg₂GeO₄) corresponds to a low-pressure structural analogue for ringwoodite. The calculated defect energies indicate that the most favourable mechanisms for hydrogen incorporation are coupled either with the reduction of ferric iron or with the creation of tetrahedral vacancies. Hydrogen will go preferentially into tetrahedral vacancies, eventually leading to the formation of the hydrogarnet defect, before associating with other negatively charged point defects. The presence of isolated hydroxyls is not expected. The same trend is observed for germanate, and thus γ-Mg₂GeO₄ could be used as a low-pressure analogue for ringwoodite in studies of water-related defects and their effect on physical properties.     

Comparison of the raman microprobe spectra of (Mg,Fe)2SiO4 and Mg2GeO4 with olivine and spinel structures

Raman microprobe (RMP) spectra were produced for each of the olivine and spinel structured phases of Mg₂GeO₄ and (Mg, Fe)₂SiO₄. The assembled data show that bands due to the tetrahedra in silicate and germanate olivines shift in a way that indicates a dominant mass effect. This correspondence is difficult to make in spinels due to differences in structural type. Differences in Fe/Mg content of olivine shift the tetrahedral vibration bands only slightly, but their linear shifts could be used to indicate the composition of the phase.     

High-pressure Raman spectroscopic study of Mn2GeO4, Ca2GeO4, Ca2SiO4, and CaMgGeO4 olivines

We present a Raman spectroscopic study of the structural modifications of several olivines at high pressures and ambient temperature. At high pressures, the following modifications in the Raman spectra are observed: 1)?in Mn₂GeO₄, between 6.7 and 8.6?GPa the appearance of weak bands at 560 and 860?cm^?1; between 10.6 and 23?GPa, the progressive replacement of the olivine spectrum by the spectrum of a crystalline high pressure phase; upon decompression, the inverse sequence of transformations is observed with some hysteresis in the transformation pressures; this sequence may be interpreted as the progressive transformation of the olivine to a spinelloid where Ge tetrahedra are polymerized, and then to a partially inverse spinel; 2)?in Ca₂SiO₄, the olivine transforms to larnite between 1.9 and 2.1?GPa; larnite is observed up to the maximum pressure of 24?GPa and it can partially back-transform to olivine during decompression; 3)?in Ca₂GeO₄, the olivine transforms to a new structure between 6.8 and 8?GPa; the vibrational frequencies of the new phase suggest that the phase transition involves an increase of the Ca coordination number and that Ge tetrahedra are isolated; this high pressure phase is observed up to the maximum pressure of 11?GPa; during decompression, it transforms to a disordered phase below 5?GPa; 4)?in CaMgGeO₄, no significant modification of the olivine spectrum is observed up to 15?GPa; between 16 and 26?GPa, broadening of some peaks and the appearance of a weak broad feature at 700–900?cm^?1 suggests a progressive amorphization of the structure; near 27?GPa, amorphization is complete and an amorphous phase is quenched down to ambient pressure; this unique behaviour is interpreted as the result of the incompatibilities in the high pressure behaviour of the Ca and Mg sublattices in the olivine structure.     

Dislocations associated with optical features in naturally-deformed olivine

Transmission electron microscopy has been used for the direct observation of dislocations in naturally-deformed olivine. The dislocations are arranged in arrays forming low-angle sub-boundaries which have been identified with features observed in the optical microscope. Comparison of this dislocation substructure with that observed in olivine, and in metals, experimentally deformed under various conditions, suggests that the deformation in nature has occurred by creep. Possible mechanisms of creep, involving the cooperative glide and climb of dislocations, are discussed.     

Dislocation recovery in fine-grained polycrystalline olivine

The rate of static dislocation recovery in Fo₉₀ olivine has been studied under conditions of high temperature and controlled atmosphere in compressively deformed polycrystals hot-pressed from synthetic (sol–gel) and natural (San Carlos) precursor powders. The sol–gel olivine, containing a small fraction of orthopyroxene, was deformed to a final strain of 19% with a maximum differential stress of 266 MPa whereas the San Carlos specimen was deformed to 15% strain and 260 MPa differential stress. Small samples cut from these deformed materials were annealed under high-temperature, controlled atmosphere conditions, for different durations to allow partial recovery of the dislocation sub-structures. Oxidative-decoration of the microstructural features, followed by backscattered electron imaging at 5 kV and image analysis, was used to determine dislocation density. The variation of dislocation density ρ with time t at absolute temperature T was fitted to a second-order rate equation, in integral form, 1/ρ(t) − 1/ρ(0) = kt with k = k ₀ exp(−E _a/RT). The activation energy E _a of the recovery process is 240 ± 43 and 355 ± 81 kJ mol⁻¹ for sol–gel and San Carlos olivine polycrystals, respectively. The measured rates are one to two orders of magnitude lower than those reported in previous studies on natural single crystal olivine. The difference may be explained by several factors such as high dislocation densities measurable from large areas at high magnification for the SEM and the technique used to estimate dislocation densities. Comparison between fine-grained sol–gel olivine and the coarser-grained San Carlos olivine aggregate did not indicate that grain boundaries play an important role in dislocation recovery, but the absence of grain boundaries might also have contributed to the high dislocation recovery rates previously measured for single crystals.     

Anisotropic growth in the olivine-spinel transformation of Mg2GeO4 under nonhydrostatic stress

The olivine-spinel phase transformation in Mg₂GeO₄ does not occur by a martensitic mechanism. The evidence, from samples transformed in a Griggs-type solid medium deformation apparatus, are:

1. (1) lack of microstructural features in the olivine phase which can be specifically associated with a martensitic mechanism

2. (2) the orientation relationship between the two phases that is predicted by the martensitic mechanism does not occur nor is there any apparent consistency of relative orientations

3. (3) application of a differential stress to the transforming sample resulted in an anisotropic growth rate for the spinel phase indicating that growth was externally controlled rather than crystallographically controlled.

Anisotropic growth of the spinel phase results in elongation of the residual olivine phase grains in the plane normal to the direction of maximum principal compressive stress. A velocity ratio of 1.7_−0.7^+5.4 has been determined for the growth rate of the spinel from measurements on residual olivine grains. The interphase grain boundary in samples transformed under stress has cusp-shaped fingers of spinel with a blunt end separated by thin spikes of olivine. Samples transformed isostatically do not exhibit this feature providing further confirmation of anisotropic growth of the spinel. The preferred growth of the spinel is consistent with a theory of phase transformation under nonhydrostatic stress. The predicted spinel finger shape based on this theory is generally consistent with observed shapes except for the blunt end. The discrepancy may be due to surface energy which has not been considered here, or to local deviations of the applied macroscopic stress.     

Hydration-induced climb dissociation of dislocations in naturally deformed mantle olivine

Widely dissociated dislocations have been observed in mantle olivine from the Erro-Tobbio peridotite in N.W. Italy. Analysis of diffraction contrast in transmission electron microscopy (TEM) indicates that the dissociation reaction involves the climb dissociation on (001) and {021} planes of b=[001] unit dislocations into partial dislocations with Burgers vectors approximately equal to 1/x 011. In the most extreme case a unit dislocation dissociates into four partials which bound three planar defects. The unusually wide dissociation and the greater extent of dissociation in olivine from amphibole-bearing rocks suggests that the dissociation is related to hydration. The occurrence of fluid inclusions along the dislocations confirms that the partials and planar defects are saturated with volatiles. Analysis of possible planar defect structures in Fo₉₀ shows that; (i) the most likely partial Burgers vectors are b=0 3/11 1/4; (ii) two of the planar defects are cation-deficient and can be stabilised by segregation of H⁺ to produce (Mg, Fe) (OH)₂ layers which are iso-structural with the OH-rich interlayer of the humite group minerals; (iii) the central planar defect is formed by removing a stoichiometric olivine (002) layer so does not produce any local chemical changes. The climb dissociation provides a possible mechanism for the transformation of olivine to a humite group mineral. OH-rich interlayers may nucleate on dislocations and extend into the crystal by climb resulting in a gradual increase of (Mg, Fe) (OH)₂ content. The only addition of material required is hydrogen which can rapidly diffuse into olivine. If the dissociation is stable and occurs at high temperatures and pressures it may significantly influence the nature and kinetics of deformation mechanisms and the olivine — spinel shear transformation mechanism in hydrated olivine.     

Deformation and recrystallisation mechanisms in naturally deformed cordierite

Cordierite — (Mg,Fe)₂Al₄Si₅O₁₈ — occurs as porphyroclasts within metapelitic and metavolcanic rocks from the Kemiö-Orijärvi belt, SW Finland. After crystallisation the cordierites have been deformed at temperatures between 550–825° C and pressures of 3–5 kbar. Optical microscopy reveals the following deformation-induced microstructures: a bimodal size distribution between host, 0.3 to 4.0 mm, and recrystallised (new) grains, 0.1 to 0.5 mm; the intracrystalline defect-structures of host grains yield undulatory extinction, subgrains and some twinning. Recrystallised grains are optically strain free. Grain and subgrain boundaries are generally straight and parallel to crystallographic low-index planes. Orientation distribution diagrams for host and recrystallised grains yield similar fabric diagrams, i.e. [010] perpendicular to foliation -S-, [001] and [100] parallel to S and [001] parallel to lineation -L-. The fabric diagrams indicate that [001] (010) is the dominant slip system. Transmission electron microscopy reveals straight free dislocations, glide and climb loops, minor {130} and {110} microtwins, isolated nodal points and dislocation walls. Contrast analyses yield Burgers vector b = [001] being dominant and b = [100] subordinate. Climb loops consist of 〈c〉-dislocations that are dissociated in (001) planes, glide loops are defined by [100] [010] and [001] (100). The cordierite microstructures have been interpreted to be generated by dislocation creep. The dominant recrystallisation mechanism is thought to be subgrain rotation subsequently followed by minor grain or twin-band boundary migration.     

Electron microscopy of olivine with perfect cleavage

Electron microscopy (SEM and TEM) of unusual olivine (Fo = 88) crystals, with perfect (010) cleavages, from Chalk Hills, Salem, Tamil Nadu has been carried out. SEM studies reveal the effect of compressive stress. Microstructures by TEM showed the abundance of curved dislocations with jogs, kinks and dipoles, indicative of the dominant climb, characteristic of high temperature deformation. The stacking fault fringes observed in olivine are due to mechanical weakening caused by nonstructural chemical constituents. The evidences for this come from fiuid-microinclusions observed and higher amount of K, Na, Rb and Sr in the olivine. Hydrolytic weakening through dislocation glide motion, assisted by H₂O containing incompatible elements, may break the metal-oxygen bonds. This leads to perfect parting of crystallographic planes under deformational stress, during solid emplacement of the dunite. The easily cleavable planes are those with the largest interplanar spacings.     

Spinelloid phases in the system MgGa2O4-Mg2GeO4

Spinelloid phases have been observed and characterized by powder X-ray diffraction and high-resolution electron microscopy. Mg₃Ga₂GeO₃(III), with a narrow composition range of approximately 3 mole percent Mg₂GeO₄, is stable at atmospheric pressure up to about 1,420° C, and is isostructural with β-Mg₂SiO₄ and the spinelloid Phase III of the NiAl₂O₄-Ni₂SiO₄ system. This represents the first occurrence of a β-phase structure stable at 1 atm pressure. Above 1,420° C (1 atm) Mg₃Ga₂GeO₈ (III) decomposes reversibly into a mixture of spinel and olivine. At high pressure (around 30 kbar at 1,100° C) it transforms into another spinelloid phase, Mg₃Ga₂GeO₈ (IV), isostructural with Phase IV of the NiAl₂O₄-Ni₂SiO₄ system. In terms of crystal structures and phase relations therefore there exists a close analogy between the magnesium gallo-germanate and nickel alumino-silicate systems, the former being a lower-pressure analogue of the latter. Our investigation of a number of other pseudo-binary spinel-olivine oxide systems suggests that the formation of spinelloid phases can be associated with the inverse character of the spinel component.     

Pressure induced phase transition in Mg2GeO4 as determined by frequency dependent complex electrical resistivity measurements

The phase transition of Mg₂GeO₄ from the olivine structure (α-phase) to the spinel structure (γ-phase) was determined by complex electrical resistivity measurements in the frequency range 10 Hz up to 100 kHz. The stability fields of the α- and the γ-phase were confirmed up to 2.05 GPa in the temperature range 845° C–1400° C. Based on volume resistivity data, a decrease of about a factor of 5 was found at the α-γ phase transition. Activation energies of electrical conduction E _a at 1.05 GPa and 2.05 GPa were calculated using the volume resistivities (?) and the relaxation times (τ). The values range from 1.98 eV up to 2.78 eV. The relative dielectric permeability increases with increasing temperature. This is is due to crystal defects and charge transport mechanisms.     

The Mg2GeO4 olivine-spinel phase transition

Enthalpies and entropies of transition for the Mg₂GeO₄ olivine-spinel transformation have been determined from self-consistency analyses of Dachille and Roy's (1960), Hensen's (1977) and Shiota et al.'s (1981) phase boundary studies. When all three data sets are analyzed simultaneously,ΔH ₉₇₃ andΔS ₉₇₃ are constrained between ?14000 to ?15300 J mol^?1 and ?13.0 to ?14.1·J mol^?1 K^?1, respectively. High-temperature solution calorimetric experiments completed on both polymorpha yield a value of ?14046±1366 J mol^?1 forΔH ₉₇₃. Kieffer-type lattice vibrational models of Mg₂GeO₄ olivine and spinel based on newly-measured infrared and Raman spectra predict a value of ?13.3±0.6 J mol^?1 K^?1 forΔS ₁₀₀₀. The excellent agreement between these three independent determinations ofΔH andΔS suggests that the synthesis runs of Shiota et al. (1981) at high pressures and temperatures bracket equilibrium conditions. In addition, no configurational disorder of Mg and Ge was needed to obtain the consistent parameters quoted. The Raman spectrum and X-ray diffractogram show that little disorder, if any, is present in Mg₂GeO₄ spinel synthesized at 0.2 GPa and 973–1048 K.     

Olivine-spinel geospeedometry: Analysis of diffusion-controlled Mg-Fe2+ exchange

Subsolidus Mg-Fe²⁺ exchange between olivine and spinel is governed by Mg-Fe²⁺ interdiffusion. Incomplete exchange results in Mg-Fe²⁺ heterogeneity in both olivine and spinel, which provides information on the thermal histories of the host rocks. A composite sphere model has been developed to obtain quantitative cooling rates or heating duration from the Mg-Fe²⁺ heterogeneity. The model assumes that a spherical core of spinel and a surrounding semi-infinite spherical shell of olivine interact by diffusion-controlled exchange of Mg and Fe²⁺. The differential equations describing the model are solved numerically by finite difference approximations. The numerical solution reveals that cooling rates or heating duration can be estimated from the relationship between the grain size of spinel and temperature calculated from the chemical compositions of the core of a spinel grain and of olivine far away from it. The calculated temperature is employed in place of $\text{Mg}\text{(Mg + Fe}{}^{\mathrm{2+}}\text{)}$ at the center of spinel to obtain the absolute temperature of thermal events.This olivine-spinel geospeedometer has been applied to peridotites. gabbro, and picrites from some ophiolite complexes in Japan to estimate their cooling rates. The estimated cooling rates for the peridotites range from 10^?4 to 10^?1 °C/yr, and those for the picrites from 10³ to 10⁴ °C/yr. The geospeedometer has been extended to estimate the heating duration of lherzolite xenoliths in basalt from the Ichinomegata crater, northeast Japan. The estimated heating duration of the xenoliths is less than one day.     

Thin intergranular melt films and melt pockets in spinel peridotite xenoliths from the Rhön area (Germany): early stage of melt generation by grain boundary melting

Optical microscopy and transmission electron microscopy (TEM) on a porphyroclastic high temperature spinel peridotite from the Rhön area reveal fine, irregular glass layers and pockets along mineral interfaces, cracks in olivine, inside olivine crystals and in spongy rims of clinopyroxene. The chemical composition of the glass deviates significantly from the composition of the host basanite. Electron diffraction technique confirms the amorphous nature of the glass, thus classifying it as a former melt. Every grain or phase boundary shows amorphous intergranular glass layers of variable thickness and characteristic chemical composition with distinct chemical inhomogeneities. Olivine grain boundaries, as the most common type of interfaces, exhibit two different types of melt glasses: (1) Type I melt at olivine grain boundaries, which is characterized by low contents of SiO₂ (～37?wt%) and Al₂O₃ (～5?wt%) and elevated contents of MgO (～31?wt%) and FeO (～22?wt%), is supposed to have formed prior to or during the thermal overprint and the dynamic recrystallisation of the xenolith in the mantle. Melt inclusions inside olivine grains with an average composition of type I melt are suggested to be earlier melt droplets at olivine interfaces, overgrown by migrating olivine grain boundaries during recrystallization in the mantle prior to the uplift of the xenolith. (2) Type II melt, the most common type of melt in the xenolith, shows higher contents of SiO₂ (～48?wt%) and Al₂O₃ (～17?wt%) but lower contents of MgO (～20?wt%) and FeO (～11?wt%). The observation of different types of glass within a single xenolith indicates the development of different chemical melt equilibria at interfaces or triple junctions in the xenolith. The absence of geochemical trends in bivariate plots excludes a unifying process for the genesis of these glasses. Melt inclusions in the spongy rims of clinopyroxene are interpreted to be the product of a potassium-rich metasomatism. The formation of most amorphous intergranular melt layers and pockets at the mineral interfaces including type II melt at olivine grain boundaries is suggested to result from decompression melting during the uplift with the basalt magma. We suggest that these glasses were produced by grain boundary melting due to lattice mismatch and impurity segregation. The observed intergranular amorphous layers or melts represent the very beginning of mineral melting by grain boundary melting.     

Transport properties of olivine grain boundaries from electrical conductivity experiments

Grain boundary processes contribute significantly to electronic and ionic transports in materials within Earth’s interior. We report a novel experimental study of grain boundary conductivity in highly strained olivine aggregates that demonstrates the importance of misorientation angle between adjacent grains on aggregate transport properties. We performed electrical conductivity measurements of melt-free polycrystalline olivine (Fo₉₀) samples that had been previously deformed at 1200 °C and 0.3 GPa to shear strains up to γ?=?7.3. The electrical conductivity and anisotropy were measured at 2.8 GPa over the temperature range 700–1400 °C. We observed that (1) the electrical conductivity of samples with a small grain size (3–6 µm) and strong crystallographic preferred orientation produced by dynamic recrystallization during large-strain shear deformation is a factor of 10 or more larger than that measured on coarse-grained samples, (2) the sample deformed to the highest strain is the most conductive even though it does not have the smallest grain size, and (3) conductivity is up to a factor of ~?4 larger in the direction of shear than normal to the shear plane. Based on these results combined with electrical conductivity data for coarse-grained, polycrystalline olivine and for single crystals, we propose that the electrical conductivity of our fine-grained samples is dominated by grain boundary paths. In addition, the electrical anisotropy results from preferential alignment of higher-conductivity grain boundaries associated with the development of a strong crystallographic preferred orientation of the grains.     

Combining ECCI and FIB milling techniques to prepare site-specific TEM samples for crystal defect analysis of deformed minerals at high pressure

Dislocation microstructures in experimentally deformed single-crystal pyrope-rich garnet, (Mg,Fe)₃(Al,Cr)₃Si₃O₁₂, and polycrystalline forsterite, Mg₂SiO₄, were investigated by using electron channeling contrast imaging (ECCI) and transmission electron microscopy (TEM) combined with a focused ion beam (FIB)-microsampling. In the orientation-optimized ECCI method, we successfully observed individual dislocations across subgrain boundaries in a low-atomic-number mineral, pyrope-rich garnet (averaged Z-numbers, AZs ～ 10). Dislocations in a deformed forsterite (iron-free olivine) were also visible in the ECCI. In the ECCI on the single-crystal garnet, deformation bands consisting of dislocations, unusual contrasts in stripes and inhomogeneous distributions of sub-micrometer-sized pores were found. Further site-specific TEM observation on the deformation band revealed a high density of partial dislocations and stacking fault ribbons. The site-specific characterizations from ECCI to TEM, with assistance of FIB, can provide a new approach to investigate dislocation microstructures of deformed materials at high pressure and high temperature.