Melt Escape Research Articles

Melt segregation rates in migmatites: review and critique of common approaches

Abstract The rate of melt segregation in regional migmatite terrains can be estimated by various lines of research. Firstly, the segregation rate of a melt batch, with a volume below the melt percolation threshold, cannot exceed the melt production rate and is therefore limited by the heating rate derived from geothermometry and geochronology (method A). Other estimates come from physical models for melt percolation (method B) and from the degree of (dis)equilibrium reached between melt and source rocks (method C). The first method is restricted by the current time resolution of isotopic techniques. Results from the second and third approaches depend heavily on assumed values of melt viscosity and other parameters (B); on the correct recognition of (dis)equilibrium trace element distributions (C); and on the migmatization model used (B and C). The validity of method C is undermined by the mathematical equivalence of trace element models for five different scenarios: (1) disequilibrium melting (with or without melt escape) followed by in situ crystallization of non-segregated melt; (2) equilibrium melting, followed by equilibrium crystallization and major melt escape; (3) disequilibrium melting, followed by equilibrium crystallization and minor melt escape; (4) pervasive retrograde re-equilibration; and (5) subsolidus differentiation. Hence, trace element data in support of model 1 (implying fast melt segregation rates) are equally consistent with models 2 to 4 (implying slow melt segregation rates), and even with melt-absent model 5. The level of trace element saturation reached during accessory phase dissolution into melt may not provide better answers, as it is largely controlled by textural constraints, e.g. shielding of accessories by porphyroblasts. We conclude that the way forward is to directly couple microtextural and microgeochemical information with time constraints. This requires high-resolution (space and time) geochronology, possibly with more advanced methods than presently available.

Transformation of two‐pyroxene hornblende granulite to garnet granulite involving simultaneous melting and fracturing of the lower crust, Fiordland, New Zealand

AbstractGranulite facies gabbroic and dioritic gneisses in the Pembroke Valley, Milford Sound, New Zealand, are cut by vertical and planar garnet reaction zones in rectilinear patterns. In gabbroic gneiss, narrow dykes of anorthositic leucosome are surrounded by fine‐grained garnet granulite that replaced the host two‐pyroxene hornblende granulite at conditions of 750 °C and 14 kbar. Major and trace element whole‐rock geochemical data indicate that recrystallization was mostly isochemical. The anorthositic veins cut contacts between gabbroic gneiss and dioritic gneiss, but change in morphology at the contacts, from the anorthositic vein surrounded by a garnet granulite reaction zone in the gabbroic gneiss, to zones with a septum of coarse‐grained garnet surrounded by anorthositic leucosome in the dioritic gneiss. The dioritic gneiss also contains isolated garnet grains enclosed by leucosome, and short planar trains of garnet grains linked by leucosome. Partial melting of the dioritic gneiss, mostly controlled by hornblende breakdown at water‐undersaturated conditions, is inferred to have generated the leucosomes. The form of the leucosomes is consistent with melt segregation and transport aided by fracture propagation; limited retrogression suggests considerable melt escape. Dyking and melt escape from the dioritic gneiss are inferred to have propagated fractures into the gabbroic gneiss. The migrating melt scavenged water from the surrounding gabbroic gneiss and induced the limited replacement by garnet granulite.

A granular flow theory for the deformation of partially molten rock

A model is developed for deformation of partially molten rock in which the primary mechanism is the relative movement of grains and the strain rate is controlled by the rate at which the interferences to this movement are accommodated by local melting. It is concluded that, in multicomponent rocks with moderate melt fractions where this model is applicable, the strain rate is generally determined by the rate of diffusion of the grain components in the melt, with linear dependence on deviatoric stress, quadratic dependence on melt fraction, and inverse quadratic dependence on grain size. Where melt is free to escape, the strain rate is somewhat higher than for a closed system but the melt escape rate is more strongly dependent on the pressure differential between grains and melt (effective pressure) than on the applied principal stress difference. Unless the stress difference is large compared with the effective pressure, any substantial increase in rate of melt release due to deformation would have to come from other factors such as increase in permeability due to microfracturing.

Melt segregation in the continental crust: distribution and movement of melt in anatectic rocks

AbstractThe grain‐ and outcrop‐scale distribution of melt has been mapped in anatectic rocks from regional and contact metamorphic environments and used to infer melt movement paths. At the grain scale, anatectic melt is pervasively distributed in the grain boundaries and in small pools; consequently, most melt is located parallel to the principal fabric in the rock, typically a foliation. Short, branched arrays of linked, melt‐bearing grain boundaries connect melt‐depleted parts of the matrix to diffuse zones of melt accumulation (protoleucosomes), where magmatic flow and alignment of euhedral crystals grown from the melt developed.The distribution of melt (leucosome) and residual rocks (normally melanocratic) in outcrop provides different, but complementary, information. The residual rocks show where the melt came from, and the leucosomes preserve some of the channels through which the melt moved, or sites where it pooled. Different stages of the melt segregation process are recorded in the leucosome–melanosome arrays. Regions where melting and segregation had just begun when crystallization occurred are characterized by short arrays of thin, branching leucosomes with little melanosome. A more advanced stage of melting and segregation is marked by the development of residual rocks around extensive, branched leucosome arrays, generally oriented along the foliation or melting layer. Places where melting had stopped, or slowed down, before crystallization began are marked by a high ratio of melanosome to leucosome; because most of the melt has drained away, very few leucosomes remain to mark the melt escape path — this is common in melt‐depleted granulite terranes. Many migmatites contain abundant leucosomes oriented parallel to the foliation; mostly, these represent places where foliation planes dilated and melt drained from the matrix via the branched grain boundary and larger branched melt channel (leucosome) arrays collected. Melt collected in the foliation planes was partially, or fully, expelled later, when discordant leucosomes formed. Leucosomes (or veins) oriented at high angles to the foliation/layering formed last and commonly lack melanocratic borders; hence they were not involved in draining the matrix of the melting layer. Discordant leucosomes represent the channels through which melt flowed out of the melting layer.

Magma flow within the Tavares pluton, northeastern Brazil: Compositional and thermal convection

Crystallization coupled with gravity removal of depleted interstitial melt has long been recognized as a mechanism of magma differentiation. Similarly, heat released by synplutonic basaltic magma intrusions has long been recognized as capable of driving convection in granite chambers. Direct evidence of these processes has seldom been described in granites. In the Tavares pluton, we mapped a number of melt extraction structures from pores of a crystal-liquid mush (an effectively solid magma where crystals form an interconnected skeleton) and a variety of flow structures such as (1) meter-scale tear- or mushroom-shaped blobs representing within-chamber diapirs; (2) meter-scale ellipsoids representing frozen thermal plumes of granite, driven by heat released from disrupted diorite intrusions; and (3) ladder dikes and snail structures representing cross sections of several superposed cylindrical magma channels (possibly feeders of diapirs and plume heads). A fundamental feature of the structures in the Tavares pluton is that they are well delineated by mafic schlieren developed at active channel margins. We postulate a new model for the origin of marginal schlieren, which combines shear flow sorting and melt escape from the flowing magma into an effectively solid surrounding mush. Extraction structures (representing melt extraction from mush pores into melt pockets) and schlieren (representing regions where melt escaped into surrounding mush pores) are both favored by magmas that form an interconnected solid framework at low crystal fractions (∼50%), because these mushes are ductile and permeable. Favorable magmas are those with a high wetting angle between melt and solid (∼60°) and a propitious crystal size and shape distribution. We propose a model of compositional and thermal convection that accounts for all described structures.

Partial melting, partial melt extraction and partial back reaction in anatectic migmatites

Anatectic migmatites commonly show both prograde (entropy producing) and retrograde reactions between minerals and melt. The final textures, mineral modes and mineral chemistries are affected by four successive processes: (i) prograde partial melting and small-scale segregation into melt-rich domains and restitic domains; (ii) partial melt extraction; (iii) partial retrograde reactions (back reaction) between in situ crystallizing melt and the restite; (iv) crystallization of remaining melt at the solidus, releasing volatiles. A new model is presented which combines the four successive processes. Partial melting is assumed to affect all textural elements of a migmatite unit in a closed system. Hence, the protolith (palaeosome) is separated into restite (now mesosome)+ melt. A batch melting model is assumed with segregation of all batches except the last. The segregated, but not extracted, melt back reacts only with the adjacent portions of the mesosome, resulting in a melanosome–leucosome pair. The last unsegregated melt batch, with a local volume fraction below the melt segregation threshold, back reacts with the surrounding mesosome. Any non-reacting melt crystallizes at or near the solidus, releasing volatiles. An important consequence of this model is that melanosome–leucosome–mesosome compositions do not necessarily show linear compositional trends in a closed system. This affects liquid compositions deduced from leucosomes, mineral modes and compositions as well as mass balance in migmatites and the possible granite–migmatite connection may therefore be blurred. Allowing water fluxes into and out of the system does not seriously affect the conclusions. A simple graphical analysis suggests that texturally observable back reaction between melt and restite may occur if certain conditions are fulfilled, most importantly: (i) fluid-absent, incongruent melting; (ii) crystallization far above the solidus; (iii) incomplete melt escape. Melanosome biotite is relictic in water-saturated partial melting as well as in reactions not consuming biotite. In other cases, melanosome biotite is partly relictic and partly produced during back reaction. The ratio of retrograde versus prograde biotite will thus depend on the protolith, on the melting reaction, and on kinetic factors.

Continuous vs. discontinuous melt segregation in migmatites: insights from a cellular automaton model

A cellular automaton model is presented that simulates melt extraction from migmatites. The varying parameters are the relative amount of melt, thresholds for melt connection leading to movement and melt escape, and type of deformation (pure and/or simple shear). It is shown that when the melt percentage is about half that of the melt escape threshold, the competition between production and deformation‐assisted extraction results in a discontinuous rate of melt extraction. The process can be compared to stick‐slip motion during frictional experiments. At higher melt fraction, it is continuously extracted. Deformation assists significantly in segregating melt, particularly when a pure shear (compaction) component is added to simple shear.

Strain partitioning during partial melting and crystallizing felsic magmas

The rheology of partly molten felsic systems changes dramatically during the onset of melting and through crystallization. Solid to liquid and liquid to solid transitions are not rheologically identical, contradicting the concept of a unique rheological critical melt percentage. We use four thresholds, separating the successive stages according to the melt fraction. For the melting transition: (1) the liquid percolation threshold connects the melt pockets at about 8% of melt, thus allowing local magma displacement at the metric scale; and (2) for greater melt percentages (20–25%), a melt escape threshold corresponds to the onset of transport over large distances of the melt and part of the residual solid phase, leading ultimately to granitic bodies. For the solidification transition: (3) the first formation of a rigid framework of crystals which can accumulate stress, although the melt can still flow, corresponds to the rigid percolation threshold (55% solids); and (4) the system becomes totally locked at about 72–75% solid, which corresponds to the particle locking threshold. The present paper focuses on strain partitioning induced by the coexistence of two contrasting rheological phases between these thresholds. On a variety of scales, non-coaxial deformation is preferentially accommodated by the lower-viscosity melt which increases the vorticity in the melt phase and reinforces asymmetries in the flow pattern. Strain partitioning also increases the scale length of the weakest phase during melting and magma ascent, whereas strain partitioning decreases the scale length during crystallizing. Strain partitioning also affects the migration of melt, by raising the particle locking threshold and lowering the melt escape threshold. For three-dimensional deformations, such as transpression, the direction of melt segregation (e.g. horizontal) may differ from the direction of melt migration (e.g. vertical).

Melt segregation in migmatites

Anatectic stromatic migmatites have a symmetrical layered structure with a low ratio of thickness to length and a periodicity, features that have not been explained satisfactorily but which are related to physical processes of melt segregation. We evaluate the compaction model for segregation as it applies to migmatites and develop models for melt segregation based upon convection driven by volume change and upon advection down pressure gradients that result from applied differential stress acting on an anisotropic multilayer protolith. Compaction by gravity‐driven two‐phase flow results in asymmetric separation, yields calculated segregation times that are slow or geologically unreasonable, and yields relative volumes that are not consistent with the abundance of leucosome seen in natural migmatites. Although calculated segregation times for “very wet” granite might allow segregation by compaction, some driving force in addition to gravity is needed to cause widespread melt segregation in the crust. For segregation at low volume fraction of melt, we develop a two‐dimensional two‐phase (matrix and fluid) viscous flow cell model (length much greater than thickness) with phase changes which predicts either expansion or contraction convection (depending on whether volume change, ΔVr, is positive or negative). Convection leads to melt segregation to form a stromatic structure in a geologically reasonable timescale. At moderate volume fraction of melt, segregation may occur by filter pressing in compositionally layered rocks in response to applied differential stress. Melt migration is by porous medium flow driven by differences in mean normal stress between the layers, as a consequence of differences in rheology, and shear‐enhanced matrix collapse. Calculated segregation rates are fast, and the model yields adequate volume of leucosome. The positive ΔVr for volatile phase‐absent melting reactions under crustal pressures promotes melt escape, unless extensional deformation facilitates melt accumulation. If the rate of melt production exceeds the rate of melt escape, then the increase in melt pressure may lead to melt‐enhanced embrittlement and fracture, and melt may migrate out of the system. In contrast, water‐rich volatile phase‐present melt‐producing reactions have a negative ΔVr, which promotes melt retention after segregation, unless deformation generates melt escape pathways. Evolution of stromatic structure through anatectic erosion of mesosome by increasing the melt fraction in situ may lead to breakdown of the solid matrix and will lead to instability due to buoyancy, and magma may become mobile and entrain residual material (“restite”).

Source component mixing in the regions of arc magma generation

Most recent workers attribute the main features of island arc basalt geochemistry to variable contributions of at least two source components. The major source appears to be the peridotitic wedge of upper mantle overlying the subducted slab, but the nature of the second component and the processes by which the sources become mixed during genesis of arc magmas are in dispute. A metasomatic addition to the wedge resulting from devolatilization in the slab is the simplest explanation of the marked enrichment of the alkali and alkaline earth elements with respect to the rare earths in island arc basalts, together with the variably developed trends in Pb, Sr, and Nd isotopic data toward sedimentary contaminants. However, lack of the correlations between relative degrees of trace element fractionation and radiogenic isotopic ratios expected of such processes requires a more complex explanation. Alternative models that suggest that all of the characteristics of island arc basalts can be accounted for by melting of an intraoceanic, hot spot type of mantle source also face specific difficulties, particularly with regard to the strong depletions of trace high‐field‐strength elements in arc compared with hot spot magmas. A possible resolution of these specific geochemical difficulties may lie in dynamic transport processes within the wedge linked with the slab through coupled drag, and the marked depression of mantle isotherms in subduction zones. Inefficient escape of melts and subsequent repeated freezing within the overturning wedge can lead to local mineralogic and geochemical heterogeneity of the peridotite overlying the slab. Fluids released from the slab may infiltrate the heterogeneous wedge and preferentially scavenge the alkalis and alkaline earths with respect to the rare earths and high field strength elements from locally enriched portions of the wedge. Incorporation of such metasomatic fluids in renewed melting at shallower but hotter levels within the wedge can give rise to the trace element and isotopic systematics generally observed in arc basalts. Furthermore, subsequent melting of wedge‐type peridotite in nonsubduction zone environments can result in complementary enrichment of the high field strength elements compared with arcs, and in the general isotopic similarity of hot spot and arc magmas. Although it is likely that the wedge‐type peridotite in any arc is heterogeneously veined by previous inefficient melt extraction episodes, it is possible that the subduction zone environment is most conducive to the generation of veining.

Geological aspects of deformation in continental shear zones

Abstract A method of kinematic analysis of structures, microstructuresand mineral preferred orientations, initially devised in the study of peridotites, has been applied to crustal rocks bearing evidence of large strains produced in metamorphic environments. Three tectonic lineaments (Angers-Lanvaux, Montagne Noire and Maydan) were selected. They illustrate a general situation arising in continental crusts when they are deformed by ductile transcurrent fault systems. The Angers-Lanvaux structure is bilaterally symmetric; its dominant feature is the horizontal stretching lineation which is parallel to the fold axes. The foliation and slaty cleavage in the most surficial formations wrap around the axis of the whole structure. The folds in the slates away from the axis also exhibit axes parallel to the general trend, but no stretching lineations. These folds are attributed to crustal shortening in a direction normal to the ductile fault. In the Montagne Noire recumbent folds are thrusted away from the axis of the structure over at least 25 km. The metamorphism is also centered on the structure and symmetrically reduced away from it. The core of the structure is occupied by a strongly lineated orthogneiss, cut by a late intrusive granite. The Maydan axial zone displays clear evidence of partial melting at various scales within the deformed gneisses: (1) in gashes perpendicular to the stretching lineation which in these anatectic formations tends to plunge at more than 45°; (2) in bands of deformed pegmatites; and (3) possibly in granites which on the one hand intrude the surrounding formations and on the other converge with increasing deformation on the fault zone. The quartz preferred orientations and microstructures in quartzite layers from Angers indicate that the plastic flow plane and direction lie, respectively, close to the foliation and lineation, the slight departure is ascribed to a flow with a rotational shear component. All this suggests a general model for the origin of such ductile zones. The horizontal relative displacement of crustal blocks along a ductile band is responsible for its overall steeply dipping foliation and horizontal lineations. Viscous heating progressively tends to concentrate the plastic flow along its axis. It is also responsible for the development of metamorphism and of anatexis at depth; the partially melted rocks tend to rise, building at shallower depth the arched structure in the axis of the ductile zone, with a continuing flowage parallel to this axis probably now in the solid state; they can also intrude the surrounding terrain as undeformed batholiths. The folds parallel to the stretching lineation in the axial zone are explained by the fact that, due to the escape of anatectic melts, the formations at depth flow in a narrowing channel. The upwelling of the axial structure induces a compression with folding in the surrounding sedimentary formations and gravity nappe sliding away from the axis.