John G. Spray

A physical basis for the frictional melting of some rock-forming minerals

Abstract Under conditions of coseismic slip (0.1–2.0 m s−1) within the upper crust, differences in the shear yield strengths, fracture toughnesses, melting points and, to a lesser extent, thermal conductivities of rock-forming minerals result in a hierarchy of friction melting susceptibilities. In many so-called pseudotachylytes it is observed that phyllosilicates and amphiboles are preferentially consumed to form the melt phase, whilst quartz and feldspar tend to survive as clasts. This can be explained in terms of the disequilibrium “flash” melting of those mineral phases possessing the lower shear yield strengths, fracture toughnesses and thermal conductivities rather than by eutectic melting that would otherwise generate equilibrium minimum melts indicative of igneous processes. At high strain rates typical of coseismic faulting, the mechanical behaviour of minerals is primarily controlled by their fracture toughnesses, with progressive comminution during slip leading to fusion. The friction melting dependence on the mechanical properties of minerals indicates that there may be an important relationship between wall-rock lithology and earthquake magnitude if frictional melting results in fault lubrication.

Pseudotachylyte controversy: Fact or friction?

High-speed slip experiments performed on Westerly granite using friction welding apparatus reveal that comminution is an essential precursor to melting by friction. Observations of slip surfaces via analytical scanning electron microscopy (SEM) document the following sequence of events occurring in 2 s with increasing velocity (up to 2 m/s): fracture; progressive comminution; surface melting of mineral fragments; fragment-to-fragment adhesion; and, finally, production of a fragment-laden, melt-supported suspension. Explosive dehydration and melting of the epidote-group mineral allanite indicates that temperatures of at least 1000 °C were realized at the interface. This is corroborated by calculation of the temperature rise for the known operating conditions. Contrary to earlier proposals, these results show that comminution and frictional melting are complementary and not mutually exclusive processes. Depending on the velocity–shear stress–displacement relations prevailing during frictional slip, rocks produced in seismogenic zones can be predominantly comminuted wall rock or fragment-melt mixes (pseudotachylytes).

Frictional melting processes and products in geological materials: introduction and discussion

This special issue is the outcome of a theme session convened by the above authors entitled “Frictional melting processes and products in geologic materials” held at the 1990 Geological Society of America Annual Meeting in Dallas, Texas, U.S.A. The goals of the session were threefold: to bring together researchers working on various aspects of frictional melting in rock, to provide a forum for the exchange of new ideas and to encourage the production of a thematic set of papers dedicated to the topic. Of the eight papers presented here, the first two (Spray, Swanson) are primarily processand mechanism-oriented. The next (Magloughlin, Maddock, O’Hara) deal primarily with chemical aspects of host-rock-pseudotachylyte relations. The last three mainly provide descriptions of naturally occurring (Koch and Masch, Techmer et al.) and artificially produced (Kennedy and Spray) friction melts. Natural pseudotachylytes are described from strike-slip zones (Magloughlin, Swanson), thrusts and their listric counterparts (Koch and Masch, Techmer et al.) and connecting lateral ramps (O’Hara). Thus, it appears that there exists an association between pseudotachylyte generation and relatively long-lived, largedisplacement faulting and shearing. The prime objective of the meeting was to promote the integration of different approaches

Viscosity determinations of some frictionally generated silicate melts: Implications for fault zone rheology at high strain rates

Analytical scanning electron microscopy has been used to determine the major element compositions of some natural and artificial silicate glasses and their microcrystalline equivalents derived by the frictional melting of intermediate to acid protoliths. The data show that the matrices of the friction melts (which cool to form pseudotachylytes) are relatively basic and hydrous, even when their protoliths are intermediate to acid. This is because frictional fusion involves the selective comminution and nonequilibrium melting of minerals based on their individual mechanical properties and melting points, not the formation of minimum melts through equilibrium mineral interaction. This means that hydrous ferromagnesian minerals (e.g., micas and amphiboles) melt preferentially to form the liquid matrix, while feldspars and especially quartz more readily survive as clasts. Pseudotachylytes generated by frictional melting are therefore not bulk melts, and as clast-melt suspensions, they cannot be considered as simple Newtonian fluids. The calculated viscosities of the friction melts are low. For example, at 1200°C, most friction melts possess zero-shear suspension viscosities of 102–104 dPa s (1 dPa s = 1 P). This is equivalent to the viscosities of tholeiitic and alkaline basaltic magmas at the same temperature. These viscosities are maximum determinations because, as clast-melt suspensions, friction melts may undergo shear thinning and exhibit pseudoplasticity at high shear rates (i.e., during slip on a fault surface). Contrary to earlier suggestions, where the bulk melting of intermediate to acid protoliths was believed to result in the generation of viscous friction melts that could act to inhibit continued sliding, this work shows that most pseudotachylytes are partial melts possessing low viscosities. The formation of highly fluid suspensions during slip may have profound effects on the dissipation of stored strain energy in the rocks surrounding a fault. Interface lubrication could facilitate an increase in the slip rate and the rate of energy dissipation. This would be manifest as an increase in high-frequency seismic wave radiation and vibrational.

Artificial generation of pseudotachylyte using friction welding apparatus: simulation of melting on a fault plane

A 150 μm thick fused layer of rock has been produced by rotating two metadolerite core faces against each other at 3000 r.p.m. under an axial load of 330 kg for 11 s using friction welding apparatus. Scanning electron microscopy and electron microprobe analysis reveal that the melt layer comprises sub-angular to rounded porphyroclasts of clinopyroxene, feldspar and ilmenite (>20 μm diameter), derived from the host metadolerite, set within a silicate glass matrix. Thermal calculations confirm that melting occurred at the rock interface and that mean surface temperatures in excess of 1400°C were attained. The fused layer shows many textural similarities with pseudotachylyte described from fault zones. Morphologically, the fused layer consists of a series of stacks of porphyroclasts welded together by melt to form ‘build-ups’ oriented at right angles to the friction surface. There is also evidence of gouging, ploughing and plucking, as well as transfer and adhesion of material having occurred between the rock faces. The mean surface velocity attained by the metadolerite (0.24 m s−1) and duration of the experiment are comparable with velocities and rise times of typical single jerk earthquakes occurring during stick-slip seismic faulting within brittle crust (i.e. slip rates of 0.1-0.5 m s−1 for, say, 1–10s). In these respects the experiment successfully simulated frictional fusion on a fault plane in the absence of an intergranular fluid. Power dissipation during the experiment was about MW m−2, comparable only to very low values for earthquakes (e.g. 1–100 MW m−2 for displacement rates of 0.1-0.5 m s−1 at shear stresses of 100–1000 bars). This indicates that melting on fault planes during earthquakes should be commonplace. Field evidence, however, does not support this contention. Either pseudotachylyte is not being recognized in exhumed ancient seismic fault zones or melting only occurs under very special circumstances.

Impact-generated carbonate melts: evidence from the Haughton structure, Canada

Evidence is presented for the melting of dolomite-rich target rocks during formation of the 24 km diameter, 23 Ma Haughton impact structure on Devon Island in the Canadian high Arctic. Field studies and analytical scanning electron microscopy reveal that the s 200 m thick crater-fill deposit, which currently covers an V60 km 2 area in the center of the structure, comprises fragmented target rocks set within a carbonate^silicate matrix. The silicate component of the matrix consists of Si^Al^Mg-rich glass. The carbonate component is microcrystalline calcite, containing up to a few wt% Si and Al. The calcite also forms spherules and globules within the silicate glass, with which it develops microtextures indicative of liquid immiscibility. Dolomite clasts exhibit evidence of assimilation and may show calcite and rare dolomite overgrowths. Some clasts are penetrated by calcite and silicate injections. Along with the carbonate^ silicate glass textures, the presence of pigeonite and spinifex-textured diopside suggests that the matrix to the crater-fill deposit was originally molten and was rapidly cooled. This indicates that the impact event that generated Haughton caused fusion of the predominantly dolomitic target rocks. It appears that the Ca^Mg component of the dolomite may have dissociated, with Mg entering the silicate melt phase, while the Ca component formed a CaCO3-dominant melt. The silicates were derived by the fusion of Lower Paleozoic sandstones, siltstones, shales and impure dolomites. Evidence for melting is corroborated by a review of theoretical and experimental work, which shows that CaCO3 melts at s 10 GPa and s 2000 K, instead of dissociating to release CO2. This work indicates that carbonate-rich sedimentary targets may also undergo impact melting and that the volume of CO2 released into the atmosphere during such events may be considerably less than previously estimated. fl 2001 Elsevier Science B.V. All rights reserved.

Generation of plagiogranite by amphibolite anatexis in oceanic shear zones

Field, geochronological, and rare earth element (REE) evidence obtained from the Fournier oceanic fragment of the Canadian Appalachians indicates that its plagiogranite component was generated by the anatexis of amphibolite and not by the fractionation of basic magma. We propose that the process occurred in two stages: first, gabbro was plastically deformed at high temperature to form dry, low-angle shear zones that subsequently evolved to amphibolite via the addition of light REE-enriched hydrothermal solutions; second, the amphibolite underwent partial melting during shear to yield a migmatite comprising bands of plagiogranite alternating with amphibolite restite. The plagiogranite locally coalesced to form pods, dikes, and lenses that injected the surrounding undeformed gabbro. We attribute the development of the Fournier plagiogranite to dynamothermal processes occurring in proximity to a spreading center due to asthenosphere-induced shear within ocean layer 3. This serves as an important illustration of the dynamic nature of metamorphism and melting that can occur in the ocean crust.

Mineralogy, provenance, and diagenesis of a potassic basaltic sandstone on Mars: CheMin X-ray diffraction of the Windjana sample (Kimberley area, Gale Crater)

Abstract The Windjana drill sample, a sandstone of the Dillinger member (Kimberley formation, Gale Crater, Mars), was analyzed by CheMin X‐ray diffraction (XRD) in the MSL Curiosity rover. From Rietveld refinements of its XRD pattern, Windjana contains the following: sanidine (21% weight, ~Or95); augite (20%); magnetite (12%); pigeonite; olivine; plagioclase; amorphous and smectitic material (~25%); and percent levels of others including ilmenite, fluorapatite, and bassanite. From mass balance on the Alpha Proton X‐ray Spectrometer (APXS) chemical analysis, the amorphous material is Fe rich with nearly no other cations—like ferrihydrite. The Windjana sample shows little alteration and was likely cemented by its magnetite and ferrihydrite. From ChemCam Laser‐Induced Breakdown Spectrometer (LIBS) chemical analyses, Windjana is representative of the Dillinger and Mount Remarkable members of the Kimberley formation. LIBS data suggest that the Kimberley sediments include at least three chemical components. The most K‐rich targets have 5.6% K2O, ~1.8 times that of Windjana, implying a sediment component with >40% sanidine, e.g., a trachyte. A second component is rich in mafic minerals, with little feldspar (like a shergottite). A third component is richer in plagioclase and in Na2O, and is likely to be basaltic. The K‐rich sediment component is consistent with APXS and ChemCam observations of K‐rich rocks elsewhere in Gale Crater. The source of this sediment component was likely volcanic. The presence of sediment from many igneous sources, in concert with Curiositys identifications of other igneous materials (e.g., mugearite), implies that the northern rim of Gale Crater exposes a diverse igneous complex, at least as diverse as that found in similar‐age terranes on Earth.

Late Eocene impact ejecta: geochemical and isotopic connections with the Popigai impact structure

Late Eocene microtektites and crystal-bearing microkrystites extracted from DSDP and ODP cores from the Atlantic, Pacific, and Indian oceans have been analyzed to address their provenance. A new analysis of Nd and Sr isotopic compositions confirms previous work and the assignment of the uppermost microtektite layer to the North American tektites, which are associated with the 35.5 Ma, 85 km diameter Chesapeake impact structure of Virginia, USA. Extensive major element and Nd and Sr isotopic analyses of the microkrystites from the lowermost layer were obtained. The melanocratic microkrystites from Sites 216 and 462 in the Indian and Pacific oceans possess major element chemistries, Sr and Nd isotopic signatures and Sm–Nd, T_(CHUR), model ages similar to those of tagamite melt rocks in the Popigai impact structure. They also possess Rb–Sr, T_(UR), model ages that are younger than the tagamite T_(CHUR) ages by up to ∼1 Ga, which require a process, as yet undefined, of Rb/Sr enrichment. These melanocratic microkrystites are consistent with a provenance from the 35.7 Ma, 100 km diameter Popigai impact structure of Siberia, Russia, while ruling out other contemporaneous structures as a source. Melanocratic microkrystites from other sites and leucocratic microkrystites from all sites possess a wide range of isotopic compositions (ϵ(^(143)Nd) values of −16 to −27.7 and ϵ(^(87)Sr) values of 4.1–354.0), making the association with Popigai tagamites less clear. These microkrystites may have been derived by the melting of target rocks of mixed composition, which were ejected without homogenization. Dark glass and felsic inclusions extracted from Popigai tagamites possess ϵ(^(143)Nd) and ϵ(^(87)Sr) values of −26.7 to −27.8 and 374.7 and 432.4, respectively, and T_(CHUR) and T_(UR) model ages of 1640–1870 Ma and 240–1830 Ma, respectively, which require the preservation of initially present heterogeneity in the source materials. The leucocratic microkrystites possess diverse isotopic compositions that may reflect the melting of supra-basement sedimentary rocks from Popigai, or early basement melts that were ejected prior to homogenization of the Popigai tagamites. The ejection of melt rocks with chemistries consistent with a basement provenance, rather than the surface ∼1 km of sedimentary cover rocks, atypically indicates a non-surficial source to some of the ejecta. Microkrystites from two adjacent biozones possess statistically indistinguishable major element compositions, suggesting they have a single source. The occurrence of microkrystites derived from a single impact event, but in different biozones, can be explained by: (1) diachronous biozone boundaries; (2) post-accumulation sedimentary reworking; or (3) erroneous biozonation.

A U/Pb age for the Shetland Islands oceanic fragment, Scottish Caledonides: evidence from anatectic plagiogranites in ‘layer 3’ shear zones

High precision U/Pb data obtained from zircons extracted from plagiogranite within the gabbro unit of the Shetland Islands oceanic fragment of northeast Scotland yield an age of 492 ± 3 Ma. Field relations indicate that the plagiogranites were generated by the partial melting of amphibolitized gabbros within high-temperature shear zones formed due to crustal deformation and fluid infiltration occurring in proximity to a spreading centre. The U/Pb data therefore constrain the crystallization age of the Shetland complex. This age is similar to U/Pb ages obtained from the Leka (497±2 Ma), Karmoy (493 +7 -4 Ma) and Gulfjellet (489±3 Ma) oceanic fragments of the Norwegian Caledonides, and the Pipestone Pond (494 3 -2 Ma) and Betts Cove (489 3 -2 Ma) oceanic fragments of the Canadian Appalachians.