Search

Mechanisms driving the long-term dynamics of plate interfaces remain poorly-constrained. To date, the rheology of the crust is considered to be controlled by solid-state diffusion processes such as crystal plastic deformation (dislocation creep). Yet, most minerals formed at high-pressure conditions are mechanically very strong (garnet, omphacite, glaucophane, zoisite, kyanite) and can only be deformed plastically at unrealistically high stresses or temperatures. A growing number of studies point to the crucial role of fluid-rock interactions and mineral transformations in the development of crustal shear zones of low viscosity. The rock weakening is interpreted as being induced by dissolution and precipitation processes at grains boundaries in chemical disequilibrium. Here, we tackle the eclogite rheology conundrum by performing the first deformation experiments at high-pressure conditions (> 2 GPa) on a two-phase aggregate representative of the lower crust.Shear experiments were performed in a new generation of Griggs-type apparatus (Univ. Orléans) at 850°C, 2.1 GPa and a shear strain rate of 10⁻6 s⁻¹. The starting material consists of mixed powders of plagioclase and clinopyroxene separated from an undeformed gabbro (Kågen, Norway) and hot-pressed with a grain size lower than 100 µm. Experiments have been conducted with 0.2% added water.Mechanical data indicate that the samples are first very strong with a peak differential stress between 1.0 and 1.4 GPa. Then, a significant weakening is observed with a stress decrease of 0.5 GPa. The high-strain samples are characterized by a strain gradient and a reaction gradient, both increasing toward the center of the shear zone. The nucleation of new phases leads to a drastic grain size reduction and phase mixing. The intensities of both are positively correlated with the strain intensity. The nature, distribution and fabric of the reaction products vary also progressively with strain intensity. At the peak stress, the reaction products are restricted to grain boundaries where they form corona structures, while in the high-strain samples, they occur throughout the sample replacing most of the starting material. The primary plagioclase and clinopyroxene grains show incipient dynamic recrystallization, whereas reaction products never do. The nano-porosity reported in the samples attests to the presence of free-fluid phase along the reactive grain boundaries, despite the high-pressure conditions. This nano-porosity requires grain boundary sliding (GBS) processes to form, as indicated by the spatially associated quadrupole junctions.Our results show that strain at eclogite-facies conditions is preferentially localized by GBS-accommodated dissolution and precipitation creep in reactive zones. We suggest that this dominant deformation process take place in rock at chemical disequilibrium in the presence of a free-fluid phase. Therefore, deformation along deep plate interfaces should be initiated and governed by transient and local transformation weakening, allowing long-term deformation at far lower stresses than dislocation creep.

Grain size is a critical parameter for viscous deformation processes and controls the rheological behaviour (weakening and strain localizations in mylonites) of polycrystalline aggregates. Recently, Ghosh et al., (2022) used Tana quartzite (~200 μm) to develop a new grain-size-insensitive flow law (stress exponent, n = 2; activation energy, Q = 110 kj/mol). The n = 2 is interpreted to indicate the serial operation of grain interior (dislocation glide) and grain boundary (dissolution-precipitation, DPC + Grain boundary sliding, GBS) accommodation processes. It is shown that all the previous results obtained from coarse-grained (∼100 - 200 μm) quartz aggregates (high total strain, significant recrystallization) plot within a factor of ∼5 times the strain rate predicted by this flow law. Moreover, a number of earlier studies with fine-grained (3.6 to 12 μm) quartz reported an n-value range of 1.7 to 2.5 (Kronenberg and Tullis 1984; Fukuda et al., 2018), similar to the values obtained from coarse-grained samples. A deviation from the grain-size-insensitive creep-dominated process is expected at those grain-size ranges and poses the question: how much weakening is possible by grain-size-reduction within the continental crust?Using a Griggs-type apparatus, we deformed fine-grained (~ 3-4 μm) as-is and 0.1 wt.% H2O-added novaculites, representing a fully recrystallized shear zone material, at the same condition as coarse-grained Tana quartzite (~200 μm). The as-is novaculite is ~3.5 times stronger than the H2O-added novaculite. A similar strength difference (~4 times) was also observed between the as-is and H2O-added Tana quartzite. The H2O-added novaculite is ~4 times weaker than the H2O-added Tana. The as-is novaculite is ~4.5 times weaker than the as-is Tana quartzite. Thus, the addition of 0.1 wt.% H2O causes a similar order of weakening as the smaller grain size; the as-is novaculite has a similar strength as the data predicted from the H2O-added Tana flow law. The maximum strength difference (more than an order of magnitude) arises between the end-member samples i.e., the as-is Tana and H2O-added novaculite, which indicate the combined effect of grain size and H2O. The n (1.91 to 2.19) and Q (118 kj/mol) values measured from novaculites are similar to the Tana and the Tana flow law can express all the novaculite data within the error associated with the Griggs-type apparatus, only by manipulating the A-value of the flow law. Therefore, the n and Q values are largely grain-size-insensitive, while the strength differences due to grain size and added-H2O can be expressed by varying the A-value. Further, the similarity in the n and Q values indicates that the same deformation mechanisms (i.e., dislocation creep accommodated by grain boundary processes) operate in both materials. Microstructural analysis of the novaculite reveals the development of a core-shell structure of individual grains due to grain boundary migration (GBM). A combination of deformation mechanisms (dislocation glide, GBM, GBS) are mutually dependent and interact to achieve the bulk sample strain. The variations in A-value may reflect the efficiency of the GBS mechanism depending on grain size and H2O content. Finally, the implications of these results on crustal strength will be discussed.

The production of micro-pores during viscous creep is a driving mechanism for fluid circulation in deep environments. However, strain-induced cracking in nature is nowadays attributed to grain boundary sliding (GBS), restricting this process to fine-grained ductile shear zones where rocks deform by diffusion creep. Here we give natural evidence of micro-cracking induced by dislocation creep, which is by far the dominant deformation mechanism in lithospheric rocks. Focusing on pure quartz shear bands across the Naxos western granite (Aegean Sea, Greece), we first document sub-micron pores that arise at grain and sub-grain boundaries. Their shape and location emphasize sub-grain rotation as a source of cracking. We then confirm that quartz is dominated by dislocation creep with evidence of a moderate to strong lattice preferred orientation (LPO) and many sub-grain boundaries, including at the margin of the pluton where the brittle/ductile transition was reached. These features coincide with (1) quartz grains located as inclusion into quartz porphyroclasts and (2) a dependency of the LPO strength on grain size. Our findings suggest that creeping cavities act as pumping sites for fluid to penetrate the crystal lattice and nucleate randomly oriented grains along sub-grain boundaries, accounting for (1) shear localization by enhancing hydrolytic weakening and (2) rock embrittlement through growth and interlinkage of cavities where phase nucleation is limited.

Quartz rheology in the presence of H2O is crucial for modelling (numerical and geophysical) the deformation behavior of the continental crust and gives important insights into crustal strength. Experimental studies in the past have determined stress exponent (n) values for flow law between ≤ 2 to 4, while the values for activation energy (Q) vary from ~120 to 242 kJ/mol. Here, we investigated the quartz rheology under high-pressure and high-temperature conditions, using a new generation hydraulically-driven Griggs-type apparatus. In order to develop a robust flow law for quartzite, we performed constant-load coaxial deformation experiments of natural coarse-grained (~ 200 μm) high purity (> 99 % SiO2) quartzite from the Tana quarry (Norway). Our creep tests were carried out at 750 to 900 °C on the as-is (no added H2O) and 0.1 wt.% of H2O added samples under 1 GPa of confining pressure. In contrast to earlier strain rate stepping experiments, the constant-load procedure needs lower strain at each step (≤1−2%) to achieve steady-state conditions. As a consequence, there is a very low amount of recrystallization. Importantly, we can determine the Q-value independently of the stress exponent (n). Microstructures from the deformed samples were characterized using polarized light microscopy (LM), SEM-cathodoluminescence (CL), and Electron backscatter diffraction (EBSD).Our creep results for both the as-is and 0.1 wt.% H2O-added samples yield Q = 110 kJ/mol and n = 2. Our microstructural analysis suggests that the bulk sample strain is accommodated by grain-scale crystal-plasticity, i.e., dislocation glide (dominantly in prism Our new flow law predicts strain rates that are in much better agreement with the inferred natural values than the earlier flow laws. It further suggests that the strength of the continental crust considering quartz rheology is significantly lower than previously predicted.

The rheology of mafic rocks buried at high to ultra-high-pressure conditions remains enigmatic. Minerals stable at these conditions are mechanically very strong (garnet, omphacite, glaucophane, zoisite, kyanite). In the laboratory, they show plastic deformation only at very high temperature (e.g. > 1000°C for pyroxene and garnet). Yet, viscous shear zones in mafic rocks metamorphosed at amphibolite and eclogite-facies conditions are commonly reported in fossil collisional and subduction zones. These shear zones localize and accommodate large amounts of strain by weakening of the host rocks. This weakening is interpreted as being induced by a transition from grain size insensitive to grain size sensitive creep, in particular with the activation of the dissolution–precipitation creep. However, the exact interplay between deformation, mineral reaction and fluid/mass transfer remains poorly-known.We have conducted a first series of deformation experiments at eclogite-facies conditions on a 2-phase aggregate representative of mafic rocks. Shear experiments were performed in a new generation of Griggs-type apparatus (Univ. Orléans) at 850°C, and 2.1 GPa with a shear strain rate of 10⁻6 s⁻¹. The starting material consists of mixed powders with < 100 µm sized grains of plagioclase and clinopyroxene from an undeformed sample of the Kågen Gabbro in Northern Norway. Experiments have been conducted with ‘as is’ (dried at 110°C) starting material and with 0.2% added water.The mechanical data indicate that the samples are first very strong with a peak differential stress at 1.4 GPa. Then, a significant weakening is observed with a stress decrease by 0.5 GPa. The high-strain sample is characterized by a strain gradient increasing toward the center of the shear zone. Metamorphic reactions occur throughout the sample, but the high-strain areas contain considerably more reaction products than the low-strain areas. The nucleation of new phases leads to a drastic grain size reduction and phase mixing, whose intensities are positively correlated with the strain intensity. The nature, distribution and fabric of the mineral products vary also progressively with the strain intensity.In the low-strain areas, dissolution-precipitation processes mainly occur along grain boundaries: plagioclase is rimmed by zoisite and a secondary plagioclase more albitic in composition while clinopyroxene is rimmed by amphibole. In the mid-strain areas, dissolution-precipitation processes are more pervasive: amphibole and a secondary more sodic clinopyroxene occurs in pressure shadows of primary clinopyroxene, while primary plagioclase is completely replaced by a fine-grained mixture of zoisite and quartz. Reaction products show a strong shape-preferred orientation parallel to the shear direction. In the high-strain areas, dissolution-precipitation leads to the nucleation of a fine-grained mixture of garnet and secondary clinopyroxene, quartz and kyanite. Most reaction products have subhedral shape with no clear preferred orientation. Hydrous minerals (amphibole and zoisite) are not observed. Our preliminary results indicate that strain at eclogite-facies conditions is preferentially accommodated and localized by dissolution-precipitation processes. Further micro-structural and geochemical analyses are required to quantify the exact interplay between the physical and chemical processes controlling the dissolution-precipitation creep.

Mafic rocks are a key constituent of the oceanic and lower continental crust. Strain localisation and fabric development in these rocks is controlled by the active deformation mechanisms. From studies of natural rocks it has been established that strain localisation and weakening in mafic rocks is directly related to fluid availability and resultant mineral reactions. Understanding the interplay between reactions, fluid availability, and deformation aids in quantifying the stresses and rates of deformation processes. We have conducted an experimental investigation to constrain the weakening mechanisms in gabbro. Shear experiments were performed in a Griggs-type apparatus at 800-900°C, and 1.2-1.5 GPa with a shear strain rate of 10⁻⁵s⁻¹. The starting material consists of mixed powders with

The effect of composition on microstructural development and mechanical strength was tested using mica-quartz-aggregates during deformation experiments.This study used two chemically different biotite minerals mixed with quartz: (1) high F-phlogopite and (2) intermediate biotite in order to investigate the role of biotite-bearing systems for the development of shear zones and strain accommodation. Shear experiments (Griggs-type apparatus) were performed using mica (30 vol. %) and quartz (70 vol. %) assemblages at 750 and 800°C, 1000 MPa and a shear strain rate of ~10-5 s-1.Mechanical results for the F-phlogopite-bearing assemblage indicate strong samples, approximately equivalent to pure quartz samples (Richter et al., 2018), deforming at differential stresses of 764-1097 MPa). F-phlogopite flakes are preferentially oriented parallel to the main shear direction, but poorly interconnected. Most of the strain is accommodated by quartz behaving as an interconnected network. Cathodoluminescence imaging reveals that quartz recrystallizes mainly by local pressure-solution and its strength controls the overall strain accommodation.In contrast, intermediate biotite assemblages are significantly weaker and deform for lower differential stresses of 290-327 MPa, as expected for natural rocks. Biotite flakes form an interconnected network accommodating most of strain.The interconnectivity of biotite grains thus plays a major role in weakening quartz-biotite assemblages. However, at similar P-T-strain and grain size conditions, the capacity of biotite grains to interconnect may also depend on its chemical composition, particularly considering the effect of trace elements incorporation (as fluorine) on the strength of the biotite interlayer bounds (Dahl et al., 1996, Figowy et al., 2021). This led us to conclude that different types of mica, behaving differently, strongly affect strength, deformation mechanism, and microstructure of the rock due to their structure, composition and stability fields.

The mechanical properties of quartz are fundamental to control the plastic behaviour of the continental crust. Our understanding of quartz rheology is still limited in the following respects: i) the large variability of flow law parameters in the earlier literature (stress exponent n = 4 to ≤ 2 and activation energy Q = 120 to 242 kJ/mol), and ii) the difficulty to identify the rate-limiting deformation mechanism, if several mechanisms are operating simultaneously. These two issues are connected and cannot be resolved separately. The present study has carried out constant-load experiments to constrain the flow law parameters of quartz. A new generation hydraulically-driven Griggs-type apparatus has been employed, resulting in reproducible mechanical data, even at very low strain rates (10−8 to10−9 s−1; so far, closest to the natural ones). Furthermore, the Q-value in constant load experiments can be estimated without prior knowledge of the n value. Our new n (= 2) and Q values (= 110 kJ/mol) are fairly low. We calculated an A-value of 1.56 × 10−9 /MPa/sec. Microstructural analysis suggests that the bulk sample strain in our experiments is achieved by crystal plasticity, i.e., dislocation glide with minor recovery by sub-grain rotation, accompanied by grain boundary migration. Micro-cracking helps to nucleate new grains. It is inferred that strain incompatibilities induced by dislocation glide are accommodated by grain boundary processes, including dissolution-precipitation creep and grain boundary sliding. These grain boundary processes are responsible for the n-value that is lower than expected for dislocation creep ( 3). The new flow law can consistently estimate strain rates (especially at low stresses) in excellent agreement with documented natural case studies and predicts a rapid drop in strength of quartz-bearing rocks in the continental crust below a depth of ∼10 km or at a temperature of ∼300 °C and higher.

Understanding Earth’s interior dynamics, the origin and factors of which maintain the present-day plate-like behavior of the lithosphere on our planet, is one of the main goals of geosciences [...]

International audience; Quartzite samples of high purity with a grain size of ∼200 μm have been experimentally deformed by coaxial shortening in a solid medium apparatus at 900°C and at confining pressures ranging from 0.6 to 2 GPa. Most samples have been shortened by ∼30% with 0.1 wt.% added H2O. The samples deformed dominantly by crystal plasticity (dislocation creep), and there is a systematic decrease of flow stress with increasing confining pressure. Strain rate stepping tests yield stress exponents of n∼1.4. The strain determined from individual grain shapes matches that determined from bulk shortening. In addition to plastic strain, mode I cracks developed in all samples, principally in the grain boundary regions. Recrystallized material, visible through cathodoluminescence colours, forms by two mechanisms: (1) progressive subgrain rotation and (2) cracking, nucleating small new grains. After high-angle boundaries have been established, grain boundary migration takes place, and a distinction of new grains nucleation origin (subgrain rotation or cracking) is impossible. At higher pressure, there is more recrystallized material forming in the deformed samples, and it is inferred that the inverse pressure dependence of flow stress is caused by enhanced grain boundary migration at higher pressure, consistent with previous studies.

Rock deformation experiments are used to compile mechanical data sets for minerals and rocks and to study microstructure and texture development.The Griggs apparatus, a solid medium piston cylinder machine was designed about 60 years ago to investigate rock deformation mechanisms and rheology at elevated confining pressures. In a typical experiment today, the confining medium is NaCl, with confining pressures up to 3 GPa, temperatures up to 1100°C, and displacement rates between 10-8 and 10-2 ms-1 (equivalent to strain rates of 10-7 to 10-3 s-1). In axial tests, the cylindrical samples are 12 to 15 mm long with a diameter of 0.625 mm. In shearing test, split cylinder assemblies are used with 0.5 to 1 mm thick samples introduced along the 45° pre-cut. Reasonable total strains are limited to 30% axial shortening or shear strains of gamma 4. (Higher strains can be attained but are difficult to analyse mechanically. Unlike for gas rigs, torsion is not available for solid medium machines).As of now, the operational fleet of solid medium deformation apparatus comprises worldwide over 20 machines in different labs (mainly in Europe, U.S.A. and Japan), providing the scientific community with an ever-growing rheological and microstructural data base.In view of numerous developments in experimental design, as well as improvements of hardware and software for data acquisition and processing, the experimental community was recently invited to a two-day workshop, hosted by the experimental group of Orléans University. The main goal was to discuss the following points:- how to further improve the apparatus, increase its scope and improve calibrations; - how to further improve data processing, and the precision and reliability of the results;- how to maintain consistency among the labs and through time (backwards compatibility);- how ensure compatibility of results from axial and shearing experiments;- how to make the data available to the community.The poster represents a condensed report of the meeting highlighting a few issues of special interest to the structural community. Visitors to the poster are invited to share their thoughts and to give us feed-back.

The ability of water to enhance plastic deformation of a quartz aggregate has been experimentally demonstrated during the sixties (e.g. Griggs and Blacic 1965), however the processes involved are still questioned. Notably the processes combining the effect of water and pressure during the deformation are still not completely understood. Pressure strongly influences the strength of fine-grained (3.6 - 4.9 µm) wet quartz aggregates (Kronenberg and Tullis 1984), where diffusion creep operates (Fukuda et al. 2018) but its effect on coarser-grained material expected to deform only by dislocation creep is not well constrained. To re-assess the effect of pressure on quartz crystal plastic deformation, natural wet quartzite samples from the Tana quarry in northern Norway (grain size ≈ 150 µm) have been deformed using a Griggs-type apparatus at varying confining pressures (from 0.6 to 2.0 GPa). All the samples with 0.1 wt. % H2O added were shortened coaxially up to 30% strain at constant strain rate (≈10-6 s-1) and temperature (900°C).All mechanical records show that quartzite flow stresses decrease systematically with increasing pressure. These results allow to determine the strength of quartzite as a function of water fugacity, such as introduced in the flow law by Kohlstedt et al. (1995) to account for both pressure and water effects. In our case, the fugacity coefficient is m≈1 when using a stress exponent of n=2.Microstructure and image analyses of samples reveal that the bulk strain results mainly from crystal plastic deformation of original grains whereas the recrystallization processes are limited volumetrically (less than 5%) and restricted to the boundaries of original grains. Deformation is not strongly partitioned into recrystallized domains compared to flattened original grains. Optical and SEM-cathodoluminescence images revealed the presence of cracks in conjunction with recrystallization (even for high-pressure samples) and associated chemical/fluid interaction, but the cracks do not contribute significantly to the bulk strain of the samples.In order to determine the amount of water used for the deformation and the redistribution of H2O during deformation, the H2O content of the quartzite has been calculated from FTIR (Fourier Transform InfraRed spectroscopy) measurements for both, grain interiors and grain boundaries. The H2O concentrations decrease inside grains with the onset of deformation with respect to the starting material. H2O is primarily stored in the grain boundary region. There is no systematic correlation with pressure. Thus, pressure dependence of H2O weakening is not restricted to fine-grained materials at high pressure and temperature. Deformation redistributes water from the grain interiors to their grain boundaries.References:Fukuda, J., Holyoke III, C.W., and Kronenberg, A.K. (2018). J. Geophysical Res.: Solid Earth, 123(6), 4676-4696.Griggs, D. T., and Blacic J. D. (1965). Science, 147(3655), 292‑295.Kohlstedt, D. L., Evans B., and Mackwell S. J. (1995). J. Geophysical Res.: Solid Earth, 100(B9), 17587-17602.Kronenberg, A. K., and Tullis J. (1984). J. Geophysical Res.: Solid Earth, 89(B6), 4281‑4297.

Deformation of lithospheric rocks regularly localizes into high-strain shear zones that include fine-grained ultramylonites. Occurring as quasi-straight layers of intimately mixed phases that often describe sharp transitions with the host rock, these structures may channelize fluid flow[1,2] and could serve as precursors for deep earthquakes[3]. However, although intensively documented, ultramylonites originate from still unknown processes. Here I focus on a mylonitic complex that includes numerous mantle ultramylonites in the Ronda peridotite (Spain). Among them, I was able to highlight one of their precursors that I better describe as a long and straight grain boundary, along which four-grain junctions are observed with randomly oriented grains of olivine and pyroxenes. This precursor starts from a pyroxene porphyroclast and extends to an incipient, weakly undulated ultramylonite, where intimate phase mixing arises with asymmetrical grain size distribution. While the finer grain size locates on one side, describing a sharp – but continuous – transition with the host rock, the grain size gradually increases towards the other side, giving rise to a smooth transition. All phases have a very weak lattice preferred orientation (LPO) in the ultramylonite, which strongly differs from the host rock where olivine is highly deformed with evidence of high dislocation densities and a strong LPO. Altogether, these features shed light on the origin of mantle ultramylonites that I attribute to a migrating grain boundary, the sliding of which continuously produces new grains by phase nucleation, probably at the favor of transient four-grain junctions. Nucleated grains then grow and progressively detach from the precursor as it keeps on migrating depending on the dislocation densities in the host rock. Although such an unusual grain boundary remains to be understood in terms of source mechanism, these findings provide new constraints on the appearing and development of ultramylonites. [1] Fusseis, F., Regenauer-Lieb, K., Liu, J., Hough, R. M. & De Carlo, F. Creep cavitation can establish a dynamic granular fluid pump in ductile shear zones. Nature 459: 974–977 (2009)[2] Précigout, J., Prigent, C., Palasse, L. & Pochon, A. Water pumping in mantle shear zones. Nat. commun. 8: 15736, https://doi.org/10.1038/ncomms15736 (2017)[3] White, J. C. Paradoxical pseudotachylyte – Fault melt outside the seismogenic zone. J. Struct. Geol. 38: 11-20 (2012)

Since quartz is among the most abundant minerals in continental crust and one of the first to show plasticity with increasing pressure and temperature, understanding its mechanical behavior is crucial for estimates on crustal strength and modeling of geodynamic processes. Since discovery of significantly lower mechanical strength of quartz as a consequence of H2O presence in the crystal (Griggs & Blacic, 1965), remarkable amount of work has been done in order to improve knowledge about processes and mechanisms responsible for so called H2O weakening effect. As the weakening effect depends on molecular H2O, it is a disequilibrium weakening process that is difficult to incorporate into existing flaw laws.In order to estimate mechanical behavior of quartz in presence of H2O, we performed deformation experiments in solid-medium Griggs-type apparatus in coaxial setting under controlled laboratory conditions using very pure natural quartzite from Tana quarry in northern Norway. Behavior of as-is and 0.1 wt% H2O added samples was studied in 1) eight shortening experiments at 900 °C, 1 GPa and constant strain rate of 10-6 s-1 reaching 5% and 30% strain, 2) six strain rate stepping experiments covering 10-5, 10-6 and 10-7 s-1, 3) two temperature stepping experiments covering 750, 850 and 950 °C and 4) one hot-pressing experiment maintaining the starting experimental conditions for 14 hours.There is a negligible strength difference between the as-is and H2O added samples. Both H2O added and as-is strain rate steeping experiments had shown surprisingly low stress exponent, with the highest value of 2.26. Temperature stepping experiments gave activation energy values of 177 kJ/mol and 198 kJ/mol. In all studied samples, strain increases towards the sample centers exhibiting grain size decrease from initial 250 – 300 µm. Three principal deformation mechanisms contributing to the bulk strain were identified: 1) crystal plasticity of original grains manifested by flattening, undulatory extinction, and development of subgrains, 2) cracking of original grains demonstrated by fluid inclusion trails and minor grain offset and 3) dynamic recrystallization via subgrain rotation recrystallization indicated by misorientation analysis from EBSD data. FTIR spectroscopy was applied to evaluate H2O speciation, quantity and distribution. Regardless of added H2O, most of deformed original grains showed relative H2O concentration between 0 and 400 H/106Si, implying significant decrease of H2O content from original 600 to 2000 H/106Si measured in undeformed grains. Average H2O concentration in grain boundaries showed 750 H/106Si for as-is samples and 1300 H/106Si for H2O added. Plasticity is most visible in CL-images, as well as higher degree of grain fragmentation and crack density in samples with added H2O. Ubiquitous presence of fluid along the grain boundaries, demonstrated by FTIR results, may have facilitated sliding along grain boundaries which, in turn, could explain the low stress exponent derived from strain rate stepping experiments.REFERENCES:Griggs, D. T. & Blacic, J. D. (1965): Quartz: Anomalous Weakness of Synthetic Crystals. Science 147(3655):292-95.

Strength profiles through the crust and upper mantle typically show the brittle to viscous transition as a change in deformation mechanism from frictional sliding to crystal plastic (dislocation creep) mechanisms. Even though such a change may conceivably take place, experimental evidence and natural observations indicate that a transition from semi-brittle to diffusion creep mechanisms rather than dislocation creep is more common.In experiments we have carried out on granitoid and mafic rock material we can distinguish 3 main processes for the brittle to viscous transition: (1) Grain size comminution by cracking produces a sufficiently small grain size (sub-micron) to cause a switch to diffusion creep. (2) Amorphous material forms (aseismically) from mechanical wear at high stresses (high dislocation densities or high work rate) without melting. The amorphous material is observed to be weak and deforms viscously. (3) Nucleation of new minerals as a consequence of metastability of existing minerals at given P,T, fluid-conditions produces fine-grained and well-mixed aggregates causing a switch to diffusion creep as in (1).The viscously deforming part of the crust or upper mantle is not the region where most earthquakes occur, because low stresses commonly are associated with viscous deformation. However, the transitions observed in experiments described above are transformational processes the material progressively evolves over a period of time in terms of microstructure, grain size, and/or composition, i.e., they are deformation-history-dependent transitions. In other words, during the transformation, only parts of the material deform by viscous processes while others have not evolved and are still brittle (and stronger). The bulk material strength of partially transformed rock depends on the connectivity of the weaker transformed material. The weaker material causes stress concentrations at the tips of transformed zones. The coalescence of transformed zones and/or a sufficiently large amount of transformed material is expected to cause catastrophic failure and thus seismic rupture. In such a way, transformation to viscously deforming weaker material may cause seismic behavior rather than according to the conventional view, where material properties change as a result of seismic deformation first, leading to creep.

Accepted manuscript version, licensed CC BY-NC-ND 4.0. In order to study the mutual effect of deformation and mineral reactions, we have conducted shear experiments on fine-grained plagioclase-pyroxene assemblages in a Griggs-type solid-medium deformation apparatus. Experiments were performed at a constant shear strain rate of 10−5 s−1, a confining pressure of 1 GPa and temperatures of 800, 850 and 900 °C. While the peak stress of plagioclase + orthopyroxene assemblages reaches values between those of the end-member phases, the strength of polymineralic materials strongly decreases after peak stress and reaches flow stresses that stabilize far below those of the weaker phase (plagioclase). This weakening correlates with the coeval development of high-strain shear zones where new phases are preferentially produced, including new pyroxene, plagioclase and amphibole. The reaction products mostly occur as intimately mixed phases within fine-grained and interconnected shear bands, together with different compositions with respect to the starting material. This indicates that deformation significantly enhances the kinetics of mineral reactions, which in turn strongly weaken the deforming sample, here attributed to a switch to grain-size-sensitive diffusion creep through phase nucleation and grain size reduction. Such an interplay between deformation and mineral reactions may have strong implications for the initiation, development, and durability of shear zones in the lower crust.

Although fundamental to the understanding of crustal dynamics in extensional setting, the relationships between the emplacement of granitic intrusions and activity of detachments still remain very elusive. Through a multi-scale approach, we here document a continuous deformation history between the monzogranitic intrusion of Naxos and the Naxos-Paros Detachment System (Cyclades, Greece). Field observations first show an early magmatic deformation followed by solid-state, ductile and then brittle deformation when approaching the detachment zone, as evidenced by the overprinting of mylonites by cataclastes and pseudotachylites. From these observations, we define six strain facies that characterize a positive strain gradient from core to rim of the Naxos monzogranite. Based on field pictures, X-ray tomography and Electron BackScatter Diffraction (EBSD) analyses along the strain gradient, we then quantify the intensity of mineralogical fabrics in 2D and 3D and better characterize the deformation mechanisms. Our measured shape variations of the strain ellipsoid corroborate the large-scale strain gradient, showing a good correlation between qualitative and quantitative studies. In addition, EBSD data indicate that dislocation creep was predominant during cooling from more than 500°C to temperature conditions of the ductile-to-brittle transition. However, 1) a weakening of quartz lattice preferred orientation with increasing strain and 2) evidence of numerous four-grain junctions in high-strain shear bands also indicate that grain boundary sliding significantly contributed to the deformation. Although the source of grain boundary sliding remains to be constrained, it provides a consistent approach to account for strain localization in Naxos.

International audience; Strain localization in viscously deformed rocks commonly results in fine-grained shear zones where massive fluid circulation is regularly observed. Recently attributed to strain-induced pumping, this phenomenon may have major implications for the distribution of ores deposits and rock rheology. However, although grain size reduction and/or creep cavitation have been proposed as important processes, the source mechanism of fluid concentration remains unresolved, particularly at high pressure. Here we use secondary ion mass spectrometry to document the H 2 O content of fine-grained olivine across an experimental shear zone, which developed with grain size reduction during a H 2 o-saturated shear experiment at 1.2 GPa and 900 °C. Through data interpolation, the olivine matrix reveals high fluid concentrations where shear strain is localized. These concentrations far exceed the predicted amount of H 2 o that grain boundaries can contain, excluding grain size reduction as a unique source of water storage. Instead, we show that H 2 o increases per unit of grain boundary across the shear zone, suggesting that cavitation and "healing" processes compete with each other to produce a larger pore volume with increasing strain rate. This provides an alternative process for fluids to be collected where strain rate is the highest in deep shear zones. On Earth, both the crust and mantle incorporate aqueous fluids that interact with solid rock materials in many ways. Commonly referred to as fluid-rock interactions, they strongly affect rock deformation and petrogenetic processes, giving rise, for instance, to hydrolytic weakening, pressure-solution creep or metamorphic reactions , including partial melting. During strain localization, this type of interactions may involve one or several chemo-physical processes that help to channelize fluid flow in ductile shear zones where grain size is substantially reduced. When rocks deform by sub-solidus viscous creep, strain indeed partitions into fine-grained shear bands that recurrently develop in the presence of massive fluid circulation, as revealed by the enrichment of hydrous phases 1-4 in the shear zones. Although the source process of such fluid flow localization is unknown at present, it may have critical implications for rock mechanics and distribution of ore deposits in deep Earth environments 5. Primarily attributed to seismic pumping of a pre-existing fault 1 , fluid infiltration is commonly inferred using evidence for dissolution-precipitation, which suggests a long-term process rather than a co-seismic one 3,5-7. Rutter and Brodie 8 and Wark and Watson 9 proposed that high fluid contents could occur in ductile shear zones as a result of fluid permeation in response to grain size reduction; because of high pressure and high temperature, supercritical fluids may distribute along grain boundaries as a uniform boundary film 9. However, recent observations of syn-tectonic water accumulation along mantle shear bands 4 do not support such a "passive" process for attracting fluids towards ductile shear zones. Instead, they suggest that fluids are dynamically driven by the deformation itself. There is abundant documentation of micro-pores produced during deformation of fine-grained material, including within natural shear zones in the middle/lower crust 5,7,10-15 and during deformation experiments on ceramics, metals and natural rocks 16-20. Also referred to as creep cavitation, the production of these micro-pores results either from grain boundary sliding (GBS) or, in a minor extent, from Zener-Stroh cracking if dislocations interact with grain boundaries 17. In both cases, the micro-cavities arise from limitations of the material to flow, particularly when diffusive mass transfer is slow at low temperature. This led several authors to propose that

International audience; The reduction of grain size is believed to play a critical role in strain localization to form shear zones. Although many mechanisms have been proposed, the source of grain size reduction remains debated. The South Armorican Shear Zone (SASZ) is a crustal-scale strike-slip fault that deforms granitoids at mid-crustal conditions. The SASZ records the transition from protolith to ultramylonite, representative of increasing ductile shear strain. To investigate the evolution of strain localization, the different states of deformation were studied using a combination of detailed microstructural, chemical and electron backscatter diffraction analyzes. Increasing strain from protolith to ultramylonite resulted in (1) grain size reduction, (2) the development of interconnected monophase layers of mica and incipient mixed-phase zones composed of phengite-quartz ± K-feldspar in protomylonite and low-strain mylonite, and (3) the formation of fine-grained mixed-phase zones composed of K-feldspar-quartz ± phengite in high-strain mylonite and ultramylonite. We propose that the causes for interconnection of mica are the formation of cracks in the protolith combined with fluid-assisted nucleation. The latter process also plays a major role in phase mixing, as attested by the precipitation of K-feldspar at triple junctions in quartz-rich layers and in fine-grained tails of inherited K-feldspar porphyroclasts in ultramylonite. The transition from quartz-rich layers and mixed-phase zones is accompanied by a strong dispersion of the quartz lattice preferred orientation. These microstructural and textural evidences suggest that phase nucleation is the major process behind phase mixing, possibly accompanied by the action of grain-boundary sliding to open cavities during deformation. Instead of a single process, we therefore highlight a succession of weakening processes in the evolution of the SASZ, starting with the crystallization of mica as a first weakening material, and then evolving with the formation of very fine-grained mixed-phase zones made principally of feldspar, quartz and phengite.

Accepted manuscript version, licensed CC BY-NC-ND 4.0. Published version available at https://doi.org/10.1016/j.jsg.2018.05.021. Shear experiments were performed in a Griggs-type apparatus at 800 °C and 1.5 GPa, at a strain rate of 2.1 × 10−5s−1 using different starting materials: (i) Powder (grain size 6–10 μm) of dry Brazil quartz with 0.15 wt% added H2O, (ii) “dry” Brazil quartz porphyroclasts (grain size ∼100–200 μm), devoid of fluid inclusions embedded in the same fine grained powder, and (iii) “wet” porphyroclasts (grain size ∼100–200 μm), containing initially a high density of μm-scale fluid inclusions embedded in the same powder. After hot pressing, samples were deformed to large shear strains (γ∼3 to 4.5), in order for the microstructures and H2O distribution to approach some state of “equilibrium”. The H2O content and speciation in quartz were analyzed by Fourier Transform Infra-Red (FTIR) spectroscopy before and after the experiments. Mechanical peak strength is generally lower in experiments with 100% hydrated matrix, intermediate in experiments incorporating wet porphyroclasts (with a proportion of 30 or 70%) and highest in those with dry porphyroclasts. All experiments with porphyroclasts show pronounced strain weakening, and the strengths of most samples converge to similar values at large strain. Wet porphyroclasts are pervasively recrystallized during deformation, while dry porphyroclasts recrystallize only at their rims and remain weakly deformed. Recrystallization of the initially fluid-inclusion-rich porphyroclasts results in a decrease in inclusion abundance and total H2O content, while H2O content of initially dry clasts increases during deformation. H2O contents of all high strain samples converge to similar values for matrix and recrystallized grains. In samples with wet porphyroclasts, shear bands with high porosity and fluid contents develop and they host the precipitation of euhedral quartz crystals surrounded by a free-fluid phase. These high porosity sites are sinks for collecting H2O in excess of the storage capacity of the grain boundary network of the recrystallized aggregate. The H2O storage capacity of the grain boundary network is determined as a H2O-boundary-film of ∼0.7 nm thickness.

International audience; In order to address geological processes at great depths, rock deformation should ideally be tested at high pressure (> 0.5 GPa) and high temperature (> 300 °C). However, because of the low stress resolution of current solid-pressure-medium apparatuses, high-resolution measurements are today restricted to low-pressure deformation experiments in the gas-pressure-medium apparatus. A new generation of solid-medium piston-cylinder ("Griggs-type") apparatus is here described. Able to perform high-pressure deformation experiments up to 5 GPa and designed to adapt an internal load cell, such a new apparatus offers the potential to establish a technological basis for high-pressure rheology. This paper provides video-based detailed documentation of the procedure (using the "conventional" solid-salt assembly) to perform high-pressure, high-temperature experiments with the newly designed Griggs-type apparatus. A representative result of a Carrara marble sample deformed at 700 °C, 1.5 GPa and 10-5 s-1 with the new press is also given. The related stress-time curve illustrates all steps of a Griggs-type experiment, from increasing pressure and temperature to sample quenching when deformation is stopped. Together with future developments, the critical steps and limitations of the Griggs apparatus are then discussed.

The Ronda Peridotite (Spain): a natural template for seismic anisotropy in subduction wedges

International audience; For the past decades, grain size reduction leading to diffusion creep in olivine is believed to be a very important process for strain localisation in the lithospheric mantle. However, the mechanisms of grain size reduction in this regime are still poorly understood (e.g., Platt, 2015). Here we show new experimental results that document grain size reduction and material weakening during wet olivine diffusion creep. While occurring for both, mono-phase and two-phase aggregates, grain size reduction is coeval with strain localisation and local phase mixing in olivine–pyroxene aggregates. Based on evidence of fluid inclusions and cracks filled with a fine-grained phase mixture, we conclude that grain size reduces as a result of fluid-assisted nucleation that takes place in the presence of an aqueous fluid during diffusion creep. Cavitation induced by grain boundary sliding (creep cavitation) can be inferred, and may play a critical role for olivine grain size reduction. Amongst their implications for rock rheology in general, our findings highlight a key process for strain localisation in the ductile uppermost mantle.

International audience; Uplift and exhumation of vast exposures of diamond facies, subcontinental mantle peridotite in the Western Mediterranean arc are attributed to tectonic scenarios including pure extension, transpression or subduction followed by delamination-driven or rollback-driven stretching. In the Ronda peridotite (southern Spain) the strong overprint of low-pressure assemblages has precluded accurate determination of the pressure and temperature conditions for the onset of exhumation that formed the spinel tectonite and garnet-spinel mylonite domain in this massif. Here we report unequivocal petrographic evidence for the existence of prekinematic, coarse-grained garnet lherzolite assemblages from the garnet-spinel mylonite domain of the Ronda peridotite. Application of well-calibrated geothermobarometers yields prekinematic minimum equilibration conditions of 2.4-2.7 GPa and 1020-1100 degrees C, demonstrating that the Ronda peridotite equilibrated at similar to 85 km depth before shearing. We also show the existence of synkinematic garnet and spinel assemblages that overprinted garnet lherzolite assemblages at 800-900 degrees C and 1.95-2.00 GPa. The decompressional cooling path and high pressure recorded by garnet-spinel mylonites rule out their formation by near-isobaric cooling above a subduction-collision wedge or during or after the emplacement of the peridotite massif into the crust. Ronda garnet-spinel mylonites represent the vestiges of subcontinental mantle ductile shear zones formed at early stages of lithosphere extension during backarc extension in the western Mediterranean. Southward to westward retreat of the African slab during the Oligocene-Early Miocene accounts for intense backarc lithosphere extension and development of the Ronda extensional shear zone, coeval with extreme thinning of the Alboran domain overlying crust.

International audience; It is now admitted that the high strength of the subcontinental uppermost mantle controls the first order strength of the lithosphere. An incipient narrow continental rift therefore requires an important weakening in the subcontinental mantle to promote lithosphere-scale strain localisation and subsequent continental break-up. Based on the classical rheological layering of the continental lithosphere, the origin of a lithospheric mantle shear/fault zone has been attributed to the existence of a brittle uppermost mantle. However, the lack of mantle earthquakes and the absence of field occurrences in the mantle fault zone led to the idea of a ductile-related weakening mechanism, instead of brittle-related, for the incipient mantle strain localisation. In order to provide evidence for this mechanism, we investigated the microstructures and lattice preferred orientations of mantle rocks in a kilometre-scale ductile strain gradient in the Ronda Peridotites (Betics cordillera, Spain). Two main features were shown: 1) grain size reduction by dynamic recrystallisation is found to be the only relevant weakening mechanism responsible for strain localisation and 2), with increasing strain, grain size reduction is coeval with both the scattering of orthopyroxene neoblasts and the decrease of the olivine fabric strength (LPO). These features allow us to propose that grain boundary sliding (GBS) partly accommodates dynamic recrystallisation and subsequent grain size reduction. A new GBS-related experimental deformation mechanism, called dry-GBS creep, has been shown to accommodate grain size reduction during dynamic recrystallisation and to induce significant weakening at low temperatures (T < 800 ¡C). The present microstructural study demonstrates the occurrence of the grain size sensitive dry-GBS creep in natural continental peridotites and allows us to propose a new rheological model for the subcontinental mantle. During dynamic recrystallisation, the accommodation of grain size reduction by three competing deformation mechanisms, i.e., dislocation, diffusion and dry-GBS creeps, involves a grain size reduction controlled by the sole dislocation creep at high temperatures (> 800 °C), whereas dislocation creep and dry-GBS creep, are the accommodating mechanisms at low temperatures (< 800 °C). Consequently, weakening is very limited if the grain size reduction occurs at temperatures higher than 800 °C, whereas a large weakening is expected in lower temperatures. This large weakening related to GBS creep would occur at depths lower than 60 km and therefore provides an explanation for ductile strain localisation in the uppermost continental mantle, thus providing an alternative to the brittle mantle.

International audience; Much debate exists concerning the strength distribution of the continental lithosphere, how it controls lithosphere-scale strain localization and hence enables plate tectonics. No rheological model proposed to date is comprehensive enough to describe both the weakness of plate boundary and rigid-like behaviour of plate interiors. Here we show that the duality of strength of the lithosphere corresponds to different stages of microstructural evolution. Geological constraints on lithospheric strength and large strain numerical experiments reveal that the development of layers containing weak minerals and the onset of grain boundary sliding upon grain size reduction in olivine cause strain localisation and reduce strength in the crust and subcontinental mantle, respectively. The positive feedback between weakening and strain localization leads to the progressive development of weak plate boundaries while plate interiors remain

International audience; Analogue and numerical models have shown that the strength of the lithospheric mantle controls the mode of lithosphere deformation. In extension, the presence or absence of a high strength brittle mantle respectively leads to localized or distributed rifting. However, first order geophysical data question the existence of such a brittle mantle. Here we use 2-D finite-element large strain modelling to quantify the impact of a ductile localizing mantle - instead of brittle - in triggering continental rifting. As a novelty, the mantle rheology considers the effect of grain boundary sliding during strain-induced grain size reduction, which may promote a significant strength drop and subsequent strain localization at low mantle temperature (< 700-800 °C). Our results reveal that such ductile localizing mantle implies varying modes of continental rifting that mainly depend on both the amount of weakening in the ductile mantle and the strength of the lower ductile crust. A medium to strong lower crust implies coupling between the upper crust and ductile localizing mantle, yielding to narrow continental rifting. In contrast, a weak lower crust implies decoupling between the upper crust and ductile localizing mantle, giving rise to a switch from distributed faulting at incipient strain to localized faulting at large strain. Ductile strain localization in the lithospheric mantle is therefore sufficient to trigger continental rifting, although a critical amount of weakening is required. Such ductile localizing mantle provides a relevant geological and mechanical alternative to the brittle mantle. It moreover provides a wider variety of modes of upper crustal faulting that are commonly observed in nature.

International audience; Olivine fabric, or Lattice Preferred Orientation (LPO), in naturally deformed peridotite largely contributes to the seismic anisotropy of the upper mantle. LPO usually results from motion of intra-crystalline dislocations during dislocation creep. In this case, experimental and numerical data indicate that the degree of mineral alignment (fabric strength) increases with increasing finite strain. Here, we show an opposite trend suggesting that olivine fabric can also result from a different deformation mechanism. Based on documentation of olivine LPOs in peridotites of a kilometer-scale mantle shear zone in the Ronda massif (Spain), we highlight a transition from a flow-parallel [a]-axis LPO (A-type fabric) to a flow-normal [a]-axis LPO (B-type fabric). While dislocation sub-structures indicate that A-type fabric results from dislocation motion, we conclude that the B-type fabric does not originate from dislocation creep, but instead from grain boundary sliding (GBS) because: (1) dislocation sub-structures remain consistent with the A-type slip system in all samples; (2) the fabric transition from A-type to B-type correlates with decreasing fabric strength despite increasing finite strain; and (3) our observations are supported by experiments that document B-type fabric in olivine aggregates where deformation involves a component of GBS. The B-type olivine fabric has a specific signature in term of seismic anisotropy, and hence, our results may have important implications for interpreting upper mantle structures and deformation processes via seismic observations.

The Ronda peridotite supplies one of the best objects to document subcontinental mantle deformation, but its internal deformation and exhumation mechanisms remain controversial. Here we provide new structural data and numerical results that constrain the Oligocene-Miocene deformation history of the Ronda massif. We first describe a mantle shear zone in the northern massif that developed subcrustal strain localization during decompression and related partial melting. The deformation regime of this mantle shear zone evolved from penetrative NE-SW stretching to sinistral shear highlighted by discrete shear bands. We then show structural observations that document a viscous deformation in the southern massif occurring prior to the thrust-assisted emplacement of the peridotite during large decompression. Finally, we performed numerical investigations that quantify a high temperature of 980°C for the basal peridotite lens at the time of its crustal emplacement. Our numerical results constrain the timing of ductile deformation of the peridotite just before 22 Ma, probably between 30 and 22 Ma. Altogether, these features led us to conclude that the deformation and exhumation of the Ronda peridotite results from lithosphere thinning subsequently inverted in the course of the Oligocene-Miocene. Among available models, our findings support the hypothesis of peridotite exhumation by the inversion of a thinned back-arc continental lithosphere during westward slab rollback through the Alboran region.

New structural data and P-T estimates of syn-deformational assemblages within the Beni Bousera peridotites and their crustal envelope are used to explain their Alpine exhumation. The Beni Bousera peridotites occur as thin sheets within high grade crustal units of the lower Sebtides (inner Rif, Morocco) and are composed of weakly deformed spinel lherzolite in the core of the massif and garnet-spinel mylonite at the rim. The main foliation trajectories in both the peridotites and overlying crustal units show systematic rotation towards their mutual contact, indicating a kilometer-scale top to the NW shearing with a dextral component along this crust/mantle contact. Widespread top to the NW shear criteria within the crustal units overlying the peridotite support this feature. Available ages constrain the development of the main foliation in both the peridotites and crustal rocks between 25 and 20 Ma. New P-T data from the peridotites show that deformation occurs during decompression from approximate to 22 kbar, 1050 degrees C to approximate to 9-15 kbar, 800 degrees C. As a consequence, exhumation of the Beni Bousera peridotites takes place during the Oligo-Miocene lithosphere thinning in the footwall of a lithospheric extensional shear zone. The exceptional preservation of garnet within the mylonitic peridotites results from rapid cooling of the border of the massif due to the juxtaposition with colder crustal rocks along this shear zone. Uplifting of the hot mantle rocks simultaneously induces high temperature metamorphism in the overlying crustal units. These new findings allow us to reconstruct the deformation history of the Beni Bousera region and the Alboran domain in the framework of the western Mediterranean geodynamics during the last 40 Myrs.. (C) 2011 Lavoisier SAS. All rights reserved

International audience; Mechanics of the continental lithosphere require the presence of a high-strength uppermost mantle that defines the "jelly sandwich" model for lithosphere strength layering. However, in deforming regions, growing numbers of geological and geophysical data predict a sub-Moho mantle strength lower than the crustal strength, or a "crème brûlée" model. To reconcile these two opposite views of lithosphere strength layering, we account for a new olivine rheology, which could promote some weakening during dynamic grain size reduction that enhances grain boundary sliding. We performed a one-dimensional numerical model of a deforming rock in order to quantify strain localization due to this weakening rheology. Strain localization occurs at temperatures lower than 800 °C and reaches a maximum at 610 °C, increasing the strain rate from 10­15 to >10­13 s­1. These results imply the existence of a sub-Moho ductile localizing mantle on a lithosphere scale, which occurs at Moho temperatures lower than 800 °C. Also, the localizing degree of this ductile mantle increases with decreasing Moho temperatures down to 550 °C. Such a ductile localizing mantle could therefore promote large strain localization during lithosphere deformation, like the brittle mantle that is commonly assumed in the "jelly sandwich" model. Furthermore, the long-term deformation (106 yr) of the ductile localizing mantle could change the lithosphere strength layering from "jelly sandwich" to "crème brûlée" in response to the grain boundary sliding­induced weakening in mantle shear zones.