Journal of Petrology Advance Access published online on May 30, 2008
Journal of Petrology, doi:10.1093/petrology/egn028
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
On the Pseudomorphing of Melt-filled Pores During the Crystallization of Migmatites
1Department of Earth Sciences, University Of Cambridge, Downing Street, Cambridge CB2 3EQ, Uk
2Sciences De La Terre, Département Des Sciences Appliquées, Université Du Québec À Chicoutimi, Chicoutimi, Québec G7H 2B1, Canada
Received September 13, 2007; Revised typescript accepted May 1, 2008
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONPseudomorphs of melt-filled pores, recognized by their generally cuspate shape, are used as diagnostic for the former presence of partial melt. They are commonly observed in migmatites from the mid- to deep crust although they occur in the smaller pores in migmatites from shallower levels (1–2 kbar). The pseudomorphing of melt-filled pores is controlled by the kinetics of nucleation and is a consequence of the greater supersaturation required for nucleation in a small pore compared with a larger one. We examine three migmatites in detail: a contact metamorphosed cherty band from an iron formation; an Archaean regional granulite from an accretionary prism; and an amphibolite-facies sample from the roots of an Archaean mountain chain. The greater undercooling required for nucleation in progressively smaller pores is recorded by the composition of plagioclase pseudomorphs. A study of dihedral angles at the corners of pseudomorphed pores demonstrates that melt–solid textural equilibrium was probably attained only in the contact aureole. The regional granulite preserves an almost unmodified reaction-controlled melt distribution, with little evidence for either melt–solid textural equilibration or solid–solid re-equilibration, whereas the reaction-controlled melt distribution in the regional amphibolite-facies example has been modified by a partial approach to solid–solid textural equilibrium. It is not clear whether the differences in dihedral angle population are due to differences in uplift and exhumation rates or due to the presence of H2O on grain boundaries.
KEY WORDS: migmatite; microstructure; dihedral angle; crystallisation
|
INTRODUCTION |
|---|
Whereas it is relatively easy to map melt distribution in migmatites once the melt has amalgamated into centimetre-scale pockets or leucosomes (e.g. Brown, 2007
), it is not always straightforward to infer the former presence of melt in rocks that have either undergone very small degrees of partial melting or in which the melt has been almost completely expelled. The grain-scale liquid distribution in melt-poor crustal rocks has commonly been inferred from the presence of highly cuspate grains with low dihedral angles (Fig. 1a). Their shape and the low dihedral angles are reminiscent of melt-filled pores in experimental charges (e.g. Jurewicz & Watson, 1985) and it has been suggested that the cuspate grains nucleated within, and infilled, melt-filled porosity (Platten, 1982, 1983; Pattison & Harte, 1988; Harte et al., 1991; Riller et al., 1996; Holness & Clemens, 1999; Sawyer, 1999, 2001; Rosenberg & Riller, 2000; Marchildon & Brown, 2002). Such grains have been termed interserts (Rosenberg & Riller, 2000), pseudomorphs after melt (Marchildon & Brown, 2002) or melt pseudomorphs (e.g. Harte et al., 1993; Clemens & Holness, 2000; Brown, 2001; Walte et al., 2005). The process of pseudomorphing of melt-filled porosity and the conditions under which it occurs are not well understood. Why do some interstitial grains form perfect pseudomorphs of a melt-filled pore (clinopyroxene in the gabbroic example shown in Fig. 1b and c), with no overgrowth of the pore walls, whereas other examples show the (multiply saturated) liquid being replaced by a poly-phase intergrowth (clinopyroxene and plagioclase in the gabbroic example shown in Fig. 1d)?
|
The mineral forming the pseudomorph is commonly different from that of the pore walls (Fig. 1a–c). In quartzo-feldspathic crustal migmatites, it is feldspar that forms the pseudomorphs in quartz-dominated volumes, whereas in feldspar-dominated volumes it is quartz (e.g. Harte et al., 1991
; Riller et al., 1996; Rosenberg & Riller, 2000). Petrographic studies of migmatites most commonly describe quartz, alkali and plagioclase feldspar pseudomorphs (e.g. Sawyer, 2001; Marchildon & Brown, 2002), although other minerals (such as cordierite and biotite) may also form pseudomorphs. Uncommonly the pseudomorph is formed of the same mineral as one or both of the walls, although it may have a different composition (e.g. plagioclase, Sawyer, 2001). The composition of single pseudomorphs is unlikely to represent the composition of the last liquid in that pore: even if other components of the liquid solidified on the walls of the pore itself, some of the liquid must have solidified elsewhere, forming pseudomorphs themselves or overgrowths on existing grains. The formation of pseudomorphs therefore necessitates mass transport through the pore network on at least the grain scale.
If averaged over a centimetre length scale, the composition of all pseudomorphs approximates that of the final liquid. Table 1 gives four examples of melt compositions estimated from point counting melt pseudomorphs within single thin-sections of meta-sedimentary migmatites; these are all of broadly granitic composition. Overgrowth of restite grains depletes the bulk composition of the pseudomorphs in the phase that forms the dominant restite phase. The data from the two granulite-facies metagreywackes from Ashuanipi in Table 1 show a low modal proportion of plagioclase, probably because plagioclase is the major residual phase in both these rocks: it is easy to underestimate the amount of overgrowth on restitic plagioclase, especially if the overgrowths are thin or in optical continuity with their substrate.
|
Developing an understanding of the pseudomorphing of melt-filled porosity necessitates the pseudomorphs being set into the context of the microstructural development of migmatites. Migmatite microstructures themselves pose a considerable interpretative problem, requiring not only an understanding of the controls on the grain-scale distribution of melt, both during and after reaction, and an appreciation of how and why textures formed during solidification may vary, but also an understanding of how these solidification textures may be modified in the sub-solidus. We thus first set the scene by providing an overview of melt distribution in partially melted rocks, together with a discussion of the role textural equilibration may play in modifying these. A review of published observations is then used to demonstrate the joint controls on crystallization textures of cooling rate and pore size: we will show that pseudomorphs form in the smallest pores in the slowest cooled rocks.
We then present detailed analysis of partially melted rocks from a contact aureole and two regionally metamorphosed migmatite terranes using petrographic observations, conventional geochemical analysis and cathodoluminescence (CL) imaging to reveal successive stages of microstructural evolution during melting and solidification. We demonstrate how the details of pseudomorph geometry can provide information on the balance of reaction and textural equilibration during anatexis, and qualitative constraints on the rates of cooling after the metamorphic peak.
|
ANALYTICAL TECHNIQUES AND CHOICE OF SAMPLES |
|---|
Plagioclase compositions and the titanium content of quartz were determined using wavelength-dispersive electron microprobe spectrometry at the Department of Earth Sciences, University of Cambridge, using a Cameca SX100 electron probe microanalyser. For the plagioclase analyses, operating conditions were an accelerating potential of 15 keV, with a nominal beam current of 10 nA for Na, Al, Si and Ca, and 100 nA for Mg, Sr, Ti, Mn and Fe, and beam diameter of 5 µm. The quartz analyses used an accelerating voltage of 25 keV, with a nominal beam current of 300 nA and a beam diameter of 5 µm. The detection limit was 7 ppm Ti. The PAP corrections procedure was used for data processing and was run using Cameca Peak Sight software.
Recent work coupling trace element concentration with CL in quartz has demonstrated a strong link between concentration of Ti and brightness (Wark & Spear, 2005), which can further be exploited using the TitaniQ thermometer (Wark & Watson, 2006; Wiebe et al., 2007). CL and back-scatter electron (BSE) images were obtained using a Jeol JSM-820 scanning microscope fitted with a Gatan MonoCL. Polished thin-sections were scanned at a large working distance (37 mm) with an accelerating voltage of 15 kV, and a beam current of 10 nA. The BSE aperture was centred while the CL detector was off-centre—this permitted BSE and CL images to be obtained over a wide area (3 mm x 3 mm) with minimal disruption. The presence of the CL detector created some astigmatism of the beam but the effect on the final images at low magnification was not significant.
CL spectra were obtained from four quartz generations discernible by variations in brightness, using a beam current of 30 nA and a dwell time of 2 s. Spectra were recorded with the Gatan MonoCL spectrograph, with a paraboloidal mirror for light collection. The spectra are shown in Fig. 2, with the detector background omitted for clarity. The darkest, most weakly luminescent, quartz has four clearly defined peaks at 320, 430, 580 and 660 nm. All four peaks are present in more strongly luminescent quartz but the 430 nm peak is dominant. The intensity of the 430 nm peak, and thus the brightness of the quartz under CL, correlates with Ti concentration, consistent with the conclusions of Wark & Spear (2005) and Wiebe et al. (2007).
|
In a texturally equilibrated poly-mineralic rock in which grains have no preferred crystallographic orientation, dihedral angles form a narrow population with a spread about the mean determined by the extent of anisotropy of interfacial energies (Herring, 1951
; Laporte & Provost, 2000; Holness, 2006). When the microstructure is out of equilibrium, the dihedral angle population may no longer be unimodal, and the standard deviation is likely to be large (Holness et al., 2005b). Although it is possible to constrain the extent of textural disequilibrium from an analysis of apparent angles measured in two dimensions using a conventional microscope stage (Jurewicz & Jurewicz, 1986; Elliott et al., 1997), a direct picture of the population of true, three-dimensional (3-D), dihedral angles can be gained using a universal stage mounted on an optical microscope (Vernon, 1997). As the details of the population of true angles are essential to interpreting disequilibrium microstructures, we measured dihedral angles using a universal stage mounted on a J. Swift monocular optical microscope at a magnification of x320. The error on each measurement is of the order of 2°, and populations of 90–162 angles were measured for each sample. The universal stage was also used to identify and obtain a qualitative picture of the distribution of grain boundary melt films, which become difficult to see if they lie at a low angle to the plane of a conventional microscope stage. |
MELT DISTRIBUTION ON THE GRAIN SCALE |
|---|
Because of the long time scales involved it is impossible to find a regionally metamorphosed anatectic rock preserving unmodified microstructures dating from the early stages of melting: primary textures are invariably overprinted. Access to the earliest stages of reaction, during which the melt distribution and geometry are totally controlled by the kinetics of reaction, is generally possible only via laboratory experiments and by examination of natural examples of contact metamorphism in which time scales of the metamorphic event were short. These show that melting results in the development of progressively thicker melt films separating reactant grains (Fig. 3a; Butler, 1961
; Mehnert et al., 1973; Büsch et al., 1974; Paquet & François, 1980; Platten, 1982, 1983; Maddock, 1986; Holness, 1999; Cesare, 2000; Holness & Watt, 2002; Holness et al., 2005a; Acosta-Vigil et al., 2006), which may contain euhedral grains of peritectic phases (Fig. 3b; Kenah & Hollister, 1983; Vernon & Collins, 1988; Grant & Frost, 1990; Brown, 2001; Marchildon & Brown, 2002; Vernon, 2004).
|
The volume increase associated with some fluid-absent melting reactions may generate over-pressure, which can result in hydrofracture (Fig. 3c and d; Clemens & Mawer, 1992
; Connolly et al., 1997; Rushmer, 2001; Holness & Watt, 2002; Holness et al., 2005b), although micro-cracking appears to be confined to experimental charges and pyrometamorphic environments. This is perhaps because only in such extreme environments is the overstep sufficient to overcome the natural tendency of systems to buffer along reactions in P–T space.
If the melt production rate is sufficiently slow (with reaction and deformation rates commensurate with that of grain-boundary mass transport) melt-bearing systems can attain textural equilibrium. Melt geometry becomes a function of porosity or melt fraction (
), the equilibrium melt–solid dihedral angle (
) and the extent of anisotropy of interfacial energies. Equilibrium dihedral angles are controlled by the balancing of the two interfacial energies involved: the grain boundary energy between the two grains of the same phase, 
b, and the energy of the interface between the two different phases, i (Fig. 4a). For isotropic materials satisfies the relation
|
|
), but because most materials of interest to geologists have significant anisotropy of interfacial energy (e.g. Kretz, 1966; Vernon, 1968; Laporte et al., 1997), the equation above is modified to incorporate torque forces, which act to rotate interfaces to lower energy orientations (Herring, 1951; Hoffman & Cahn, 1972; Cahn & Hoffman, 1974). This stabilizes planar interfaces at pore corners (e.g. Kretz, 1966; Vernon, 1968; Laporte & Watson, 1995; Cmíral et al., 1997; Lupulescu & Watson, 1999; Laporte & Provost, 2000), and a range of possible equilibrium angles, depending on the relative orientation of the solid grains (Herring, 1951; Laporte & Provost, 2000).
|
For geologically relevant systems the equilibrium melt–solid dihedral angle is less than 60° [see Holness (2006
) for a recent review] with a standard deviation of 
12° (Laporte & Provost, 2000; Holness, 2006) so the equilibrium distribution is one of interconnected channels along three-grain junctions (Fig. 4b and c; Smith, 1964; Beeré, 1975; von Bargen & Waff, 1986). For isotropic materials these channels remain open and interconnected even for vanishingly low porosities but, because all minerals are anisotropic to some extent, there is a threshold level for interconnectivity of less than a few volume per cent (Wark & Watson, 1998; Maumus et al., 2004; Price et al., 2006; Yoshino et al., 2006).
In a system in which liquid can move freely, complete equilibration will entail melt flow (e.g. Jurewicz & Watson, 1985) until it reaches the minimum energy porosity, which ranges from 0 vol. % at >60° to 23 vol. %, or the porosity required for complete disaggregation, for = 0° (Park & Yoon, 1985; Cheadle, 1989). For the likely range of equilibrium melt–solid dihedral angles in geological systems (20–40°, Holness, 2006), the minimum energy porosity is 10–18 vol. % (Cheadle, 1989). The corollary of this is that textural equilibration in a melt-present system in which the liquid is free to move will result in the infiltration of melt along previously dry three-grain junctions (Watson, 1982; Daines & Kohlstedt, 1993; Hammouda & Laporte, 2000).
The extent of textural equilibration depends on the rate and mechanism of deformation during anatexis. Although diffusion creep does not change the melt distribution, dislocation creep can result in a significantly different melt distribution compared with the static case (Ave’Lallemant & Carter, 1970; Urai, 1983; Jin et al., 1994; Hirth & Kohlstedt, 1995; Kohlstedt & Zimmerman, 1996; Daines & Kohlstedt, 1997). Similarly, if the rates of reaction, whether solidification in igneous rocks or melting in migmatites, are faster than that of textural equilibration, melt geometries during reaction will be controlled by reaction kinetics.
Much of our understanding of the process of textural equilibration has been obtained from studies of plutonic igneous rocks (Hunter, 1987; Holness et al., 2005b; Higgins, 2006). These show that textural equilibration, or maturation, takes place in a series of stages, the earliest of which is the rotation of grain boundaries in the vicinity of three-grain junctions to achieve the equilibrium dihedral angle (Holness et al., 2005b). As a consequence of angle change, the grain boundaries in the vicinity of the pore corner develop a step-change in curvature. Because changes in mean curvature increase the energy of the interface (Beeré, 1975), the pore wall must move to attain constant mean curvature, at least in isotropic systems (e.g. Holness & Siklos, 2000). There is thus a complex interplay between the gradually changing angle at the melt–solid–solid junction and the propagation of a change in grain boundary orientation and curvature further from the junction.
Because continuous films of liquid on grain boundaries are stable only for equilibrium melt–solid dihedral angles of 0° (Smith, 1964), and equilibrium dihedral angles are always >0° for silicate systems, reaction-generated grain boundary melt films will neck down to form strings of melt-filled lenses of a shape controlled by the dihedral angle. Melt-filled, intra-grain cracks formed by hydrofracture will heal, leading to arrays of melt-filled inclusions with a shape controlled by the interfacial energies of the host crystal.
The details of establishment of the equilibrium dihedral angle at pore corners depends on the geometry of the initial reaction-controlled melt geometry and on whether the equilibration takes place in the super- or sub-solidus (or both). Thin grain boundary melt films, generated either by reaction between adjacent phases or by inter-grain fracturing, in a rock with an initially well-equilibrated solid-state microstructure will initially generate dihedral angles of 120° (Fig. 5). As the walls melt back and the melt rim becomes thicker, the dihedral angles will tend towards 180°, with a low standard deviation. Melt films propagating down two-grain junctions are akin to propagating fractures and likely to have very low dihedral angles at their tips. Melt–solid–solid dihedral angles in a melting rock in which melt distribution is dominated by grain boundary films will therefore form as many as three separate peaks (Fig. 5c). The rate at which pore junctions will progress towards melt-present equilibrium will depend on their geometry. The tips of propagating films are close to melt–solid–solid equilibrium and will not change significantly. The junctions of grain boundaries with thick melt films are farthest from equilibrium, will have the greatest driving force for microstructural change, and will therefore undergo rapid adjustment of the pore walls. The peak at 120° will migrate relatively slowly towards lower angles.
|
|
MICROSTRUCTURES FORMED DURING CRYSTALLIZATION |
|---|
Cooling rate exerts a primary control on the microstructures formed during crystallization. At one extreme we find supercooled melt (glass), whereas more slowly cooled melt crystallizes as increasingly coarse polycrystalline aggregates as the cooling rate is decreased. This can be illustrated by a consideration of quartzo-feldspathic rocks from a series of contact aureoles.
Pyrometamorphism at 120 bar around the 50 m diameter gabbro plug at Glenmore, Ardnamurchan, resulted in glassy melt rims separating quartz and feldspar (Fig. 3a and b; Butler, 1961; Holness et al., 2005a). Solidification of melt rims in the 150 bar aureole of the gabbro plug in Kinloch, Isle of Rum, and in the 600 bar aureole of the Traigh Bhan na Sgurra sill (Holness & Watt, 2002) resulted in a fine granophyric intergrowth (Fig. 6a). Solidification of melt rims in amphibolitic gneiss metamorphosed by the Rum Igneous Complex at 100–200 bar formed a coarser granophyric intergrowth in which quartz paramorphs after tridymite are visible (Fig. 6b and c; Holness & Isherwood, 2003).
|
A further reduction in cooling rate results in melt rims on quartz–feldspar grain boundaries being replaced by either an aggregate of equigranular, equant grains of feldspar and quartz (termed the string of beads texture; Holness & Isherwood, 2003
) with randomly oriented grains, or by a highly irregular grain boundary (Fig. 6e and f). The irregularity of the latter is due to the formation of extremely coarse granophyric intergrowths, with each phase having nucleated and grown on its respective side-wall of the melt rim. Which of these two textures forms is probably a consequence of the ease of nucleation during solidification. As the cooling rate decreases further, for larger or deeper contact aureoles and for regional metamorphism, the extended time scale permits the onset of melt migration, resulting in the grain-scale draining of melt from its localized source, the destruction of thick, parallel-sided, melt films such as those depicted in Fig. 6b, and segregation of melt into large pockets. This permits the identification of the second major control of crystallization microstructures: the size of the melt pocket.
In the deeper parts of the contact aureole associated with the Rum Layered Suite (Holness & Isherwood, 2003), and in that associated with the 3 kbar, 10 km diameter, Ballachulish Igneous Complex (Pattison & Harte, 1988), the largest melt pockets are filled with granophyre (Fig. 7a) or by oikocrysts with abundant inclusions (Fig. 7b; Grant & Frost, 1990; Harte et al., 1991). Thick melt films on quartz–feldspar grain boundaries are replaced by highly cuspate intergrowths or a string of beads (Fig. 6e and f), whereas the smallest pores (the thinnest melt rims, and melt pockets bounded by 3–4 grains) are generally pseudomorphed by single crystals of the volumetrically minor phase (Fig. 7c).
|
For migmatites in the mid- to deep crust, the same pattern of solidification microstructures is observed. Smaller melt pockets (typically those bounded by fewer than four grains) are partially or wholly pseudomorphed by single grains. In contrast, large melt pockets, or leucosomes, crystallize as poly-mineralic aggregates (although granophyric intergrowths are apparently confined to shallow to mid-crustal environments). We suggest that this microstructural progression, from poly-mineralic aggregates in the larger pores to monocrystalline pseudomorphs of the smaller pores, which can be observed within a single rock (and therefore at a constant cooling rate), reflects an increasing barrier to nucleation as the pore size becomes smaller.
The role of pore size in nucleation
Because an atom on the surface of a small crystal suspended in a solution is in a highly energetic condition compared with an atom of the same material on a planar surface (i.e. on the surface of an infinitely large, spherical, crystal) in contact with the same solution, it has a stronger tendency to leave the crystal (i.e. to dissolve). To maintain chemical equilibrium, to permit a correspondingly higher rate of movement of atoms from the solution onto the crystal, the small crystal needs to be in contact with a stronger solution than does a larger crystal. The concentration of solute, C, in equilibrium with a spherical crystal of radius r is given by
|
|
is the energy of the solid–liquid interface, R the gas constant, T the temperature, V the molar volume of the solid phase, and C0 is the solubility of a liquid in equilibrium with an infinitely large spherical crystal or a planar interface (Cahn, 1980; Adamson, 1990).
This means that the thermodynamics of crystallization in pores is different from that in a free fluid: most importantly, a confined fluid can become more supersaturated before crystallization begins compared with an unconfined fluid of the same composition (e.g. Bigg, 1953; Melia & Moffitt, 1964; Cahn, 1980; Scherer, 1999; Putnis & Mauthe, 2001). For liquid-bearing rock with a range of pore sizes the point at which crystals (be they interstitial grains in a plutonic igneous rock or cement in a sedimentary rock) nucleate will be determined by pore size. The first observation of this effect in rocks was reported by Putnis & Mauthe (2001), who showed that small pores in sandstone remain empty whereas nearby larger pores are filled with halite cement. A similar effect is probably implicated in the preservation of the thick grain boundary films of melt in almost completely solidified crystalline mafic nodules described by Holness et al. (2007).
|
TEXTURAL MODIFICATION IN THE SUB-SOLIDUS |
|---|
Further microstructural adjustment towards lower internal energies may occur in the sub-solidus (Fig. 5d). This process is much slower than the melt-present case because mass transfer is achieved by grain boundary diffusion rather than mass transport through a liquid. Furthermore, whereas sub-solidus textural equilibration of microstructures in monomineralic domains is rapid, as it necessitates only mass movement across grain boundaries (Hunter, 1987
), equilibration in polymineralic domains requires mass transport along grain boundaries and is therefore sluggish. It is rare to find completely equilibrated solid-state microstructures in polymineralic rocks.
Typical median angles for solid-state textural equilibrium fall in the range 100–140° (e.g. Kretz, 1966; Vernon, 1968, 1970), with standard deviations of 10–20° (Vernon, 1968, 1970). Therefore it is the pseudomorphed propagating film tips that are furthest from equilibrium and likely to migrate the fastest (Fig. 5d). The 180° angles on thick melt films will migrate to lower angles relatively slowly.
|
CONTACT-METAMORPHOSED QUARTZO-FELDSPATHIC ROCKS AT 2 KBAR |
|---|
The Lower Proterozoic Biwabik Iron Formation in Minnesota comprises a cherty and slaty banded iron formation (Morey, 1992
) and is part of the Animike Group, which was metamorphosed at 2 kbar by the Duluth Igneous Complex, an arcuate mass of troctolitic and anorthositic intrusions of Middle Proterozoic age located in the mid-continent rift of North America (Van Schmus & Hinze, 1985). Samples of cherty horizons of the Biwabik Formation were obtained from a drill core through the intrusion and into the footwall.
The cherty horizons are dominated by coarse-grained quartz, feldspar, clinopyroxene (with well-developed exsolution lamellae of Ca-poor pyroxene), magnetite and ilmenite (Fig. 8a). Coarse-grained granophyric intergrowths are common and mark centimetre-sized pockets of melt (Fig. 8b). Quartz-rich parts of the rock are dominated by rounded quartz grains separated by a network of thin single crystals of plagioclase (with minor clinopyroxene), which outline the grain boundaries and form cuspate grains at three-grain junctions. Each plagioclase grain is approximately one, or at most two, quartz grains in length. This network occupies 10 vol. % of the rock. A similar pattern is observed in relatively quartz-poor regions, in which the Fe-oxides and clinopyroxene form the network of interserts (Fig. 8a).
|
A hundred randomly chosen plagioclase–quartz–quartz dihedral angles form a unimodal population with a median of 68°, and a standard deviation of 17·3° (Fig. 9b). The median is significantly higher, and the standard deviation slightly higher, than expected for melt–quartz equilibrium (
20° and 10°, respectively; Holness, 2006). For comparison with solid-state equilibrium, plagioclase–quartz–quartz angles were measured in texturally well-equilibrated regions of an upper amphibolite-facies rock (sample EL180, which is described at length below). The median and standard deviation of a population of 100 angles are 112° and 16·0°, respectively (Fig. 9a), comparable with a mean angle of 105° (and standard deviation of 12·4°) measured in granulites from Broken Hill (Vernon, 1968), and also with a median of 117° measured by transmission electron microscopy in pelitic schists (Hiraga et al., 2002). These angles are significantly higher than the median measured in the Biwabik migmatite.
|
CL imaging reveals four distinct quartz generations (Fig. 10). Dark, weakly luminescing lobate cores are overgrown by a slightly more luminescent quartz generation, which does not completely surround the cores in many of the grains, signifying a period of quartz dissolution before the growth of the third quartz generation. The latter forms brightly luminescent, discontinuous rims around the partially resorbed grains, and is rare in the quartz-rich zones. The final quartz generation is slightly less luminescent than the brightest generation and forms continuous rims around the quartz grains. This fourth generation is that of the granophyre in the large melt pockets, and forms wide margins and irregular extensions in optical continuity with the quartz grains bounding the granophyric pockets. It forms only thin and discontinuous rims in the quartz-rich regions.
|
Having established that luminescence intensity is correlated with Ti content (Fig. 2), temperatures corresponding to each of the four quartz generations along four profiles across quartz grains were determined using the TitaniQ geothermometer of Wark & Watson (2006
). Because rutile is absent, the temperatures (shown in Fig. 10) are minimum estimates, although the relative differences between the generations are accurate if TiO2 activity remained constant (Wark & Watson, 2006). This assumption depends on the mineral assemblage remaining the same throughout the metamorphic history. Although a complete assessment of the TiO2 activity would necessitate examination of lower-grade samples from the aureole, ilmenite is common in these rocks and we believe the TiO2 activity was most probably close to unity for the duration of quartz recrystallization and growth. With these caveats, the dark, resorbed cores give a temperature of 630 ± 10°C. That of the first overgrowth is variable, with a monotonic increase to 730–800°C. The bright, discontinuous overgrowths have temperatures between 870 and 990°C, whereas the outermost rims and granophyric intergrowths record temperatures of 820–830°C. Plagioclase compositions were determined across a single thin-section (sample ELY-6B), with randomly chosen analyses within the large granophyric pocket and the thin melt pseudomorphs. The composition is shown as a function of the width of the plagioclase interserts (or the width of the granophyric pockets) in Fig. 11. For transitional regions, in which the walls of the pocket are lobate, the width is taken to be that at the widest part. The plagioclase composition varies by a few molar per cent about an average of Ab51 until the width of the pockets decreases below 40 µm. For smaller pockets the plagioclase is more albitic, reaching Ab88 for a significant proportion of the thinnest grain boundary films.
|
|
REGIONAL QUARTZO-FELDSPATHIC GRANULITES |
|---|
Ashuanipi Subprovince, Quebec
The Ashuanipi Subprovince, lying NE of the Opatica Subprovince, forms the eastern end of a 2000 km long meta-sedimentary belt that crosses the entire width of the Superior Province (Card & Ciesielski, 1986
; Percival et al., 1992) and is believed to be the remains of a single, late Archaean, accretionary prism (Percival, 1989). Sample DL96-1006A [previously described by Sawyer (2001)] is a meta-greywacke that experienced a prolonged (40–50 Myr, Guernina & Sawyer, 2003) metamorphic event that peaked in the granulite facies at about 850°C and 6–6·5 kbar (Guernina & Sawyer, 2003; Percival, 1991). It is a foliated, medium-grained, quartzo-feldspathic rock containing abundant plagioclase, biotite and orthopyroxene, and with a melt-depleted bulk composition (Guernina & Sawyer, 2003).
Melt, pseudomorphed predominantly by non-twinned K-feldspar (which appears only as pseudomorphs), with subordinate amounts of quartz and plagioclase, is concentrated in films and pools around reactant biotite grains (Fig. 12a). Quartz and plagioclase pseudomorphs tend to be more compact than the K-feldspar, with high dihedral angles against the adjacent quartz and plagioclase grains. In contrast, K-feldspar pseudomorphs form extended, cuspate films several grain diameters in length. Melt films occur mainly on biotite–quartz, biotite–plagioclase, and plagioclase–quartz grain boundaries and are associated with rounded, corroded grains of biotite, plagioclase and/or quartz (Fig. 12a and b), suggesting the reaction
|
|
). Melt films on plagioclase–plagioclase and quartz–quartz boundaries are uncommon, are generally in close proximity to biotite grains, and are always a continuation of a thick melt film on a plagioclase–quartz or biotite–quartz grain boundary. The thinnest films terminate along an inter-phase boundary (Fig. 12c and d) and isolated thin lenses may be present. Slightly thicker films are usually only a single grain width long and terminate at a three-grain junction. The thickest films are several grain diameters in length, with several melt-free grain junctions terminating at the film wall (Fig. 5a and b).
|
Quartz is generally featureless under CL, apart from a slightly brighter grain core compared with the rim, and the presence of numerous, late-stage, dark fractures (Fig. 13). The absence of the kind of detail observed in the Biwabik samples is probably because Ti variations are able to decay by diffusion during extended high-temperature episodes.
|
Dihedral angles at the junctions of the K-feldspar melt pseudomorphs with either plagioclase or quartz were measured in sample DL96-1006A, which is typical of the terrane (Sawyer, 2001
). The angle population is strongly bimodal (Fig. 9c). The bimodality is controlled by the geometry of the melt film. The terminations of the thinnest melt films have low dihedral angles, forming a well-defined peak at about 20° (examples are labelled ‘a’ in Fig. 14). The slightly thicker melt films, which terminate at three-grain junctions, have angles generally below about 60°, whereas the junctions of melt-free grain boundaries, which terminate at the thickest melt films, form a well-defined peak at 100° (examples of which are labelled ‘b’ in Fig. 14). Although this does correlate with mineralogy, as low-angle, terminating melt films almost invariably occur on plagioclase–quartz grain boundaries whereas it is predominantly quartz–quartz grain boundaries that terminate at the thickest melt films (Figs 12 and 14), we do not think that these variations are caused by interphase variations in interfacial energy. This is because equilibrium dihedral angles for melt–quartz compared with melt–plagioclase, and quartz–plagioclase–plagioclase compared with plagioclase–quartz–quartz systems are very similar (Vernon, 1968; Holness, 2006).
|
Opatica Subprovince, Quebec
The Opatica Subprovince is a 200 km wide belt of amphibolite-facies plutonic gneisses from the southeastern part of the Superior Province, Quebec. It is interpreted as the deeply eroded interior of an Archaean mountain chain and represents a section of reworked Archaean crust without major tectonic disruption (Sawyer & Benn, 1993
). The highest-grade, and later exhumed, migmatites are in the centre of the Subprovince and record maximum temperatures of 760–810°C at a pressure of 6–7 kbar (Sawyer, 1998).
Sample EL180, collected from approximately 50°33'N and 77°33'W, is a typical residual grey gneiss containing plagioclase, green hornblende and biotite, with accessory apatite and zircon [see Sawyer (1998) for a bulk composition analysis]. Titanite is abundant and forms strings of small grains within aggregates of biotite and amphibole. Orthopyroxene is absent. Biotite grains are commonly partially chloritized, and are deformed by the adjacent growth of lenses of untwinned K-feldspar, many of which are partially albitized. Both chloritization and lens growth are indicators of low-temperature migration of hydrous fluids (Holness, 2003), which therefore significantly postdate formation of the migmatites.
Solidified melt is pseudomorphed by microcline, plagioclase and quartz. Extended, fine-grained, poly-mineralic aggregates commonly occur along boundaries between larger grains (Fig. 15a), and we interpret these as solidified melt rims. The quartz and plagioclase grains tend to be equant, with 120° dihedral angles against adjacent quartz and plagioclase grains (Fig. 15b). Only the microcline grains are cuspate and elongate, with dihedral angles <120°, reminiscent of the interserts described elsewhere (Fig. 15b and c), but even the microcline pseudomorphs are relatively rounded and equant compared with those in the Ashuanipi granulite (Fig. 12).
|
Melt is not concentrated adjacent to biotite or amphibole but is evenly distributed, forming thick grain boundary films on plagioclase–plagioclase, plagioclase–quartz, plagioclase–biotite and quartz–biotite grain boundaries, together with compact, triangular pools at quartz or feldspar three-grain junctions: these are rounded with blunt apices. Thin melt films are uncommon, and intra-grain cracks are rare (only one example was observed).
The melting reaction is not as clear-cut as in the Ashuanipi rocks. However, the dearth of melt around biotite grains, the absence of peritectic phases, and the concentration of melt rims on plagioclase–quartz grain boundaries, suggest that it may have been
|
|
). The source of the hydrous fluid that drove melting is uncertain.
In a similar fashion to the Ashuanipi samples, the Opatica rocks do not reveal any detailed information in CL images. Once again, the study concentrated on dihedral angles formed at the corners of melt pores pseudomorphed by K-feldspar. The dihedral angle population is bimodal with peaks at 70° and 100° (Fig. 9d). Angles below 50° are rare. In a similar manner to DL96-1006A, the lower peak corresponds to the tips of grain boundary melt films and the corners of pores at three-grain junctions, whereas the higher peak corresponds to grain boundaries truncated by melt films.
|
THE DEVELOPMENT OF PORE PSEUDOMORPHS |
|---|
The best information about the development of pore pseudomorphs is preserved in the cherty bands in the Biwabik Formation. The four generations of quartz shown in Fig. 10 record metamorphism, anatexis and solidification of the protolith. The last episode of quartz growth, to form the granophyre, occurred at a temperature consistent with melt–solid chemical equilibrium in the Qtz–Ab–H2O system at 2 kbar (Johannes, 1978
; although he presented data only for 5 kbar). The earlier growth of highly luminescent quartz may have occurred during a period of lower H2O activity (and higher melt temperatures); the details of this remain unclear.
The quartz overgrowths on the restitic grains bounding the largest pores show that at the metamorphic peak the melt was saturated only in quartz. This melt occupied 20 vol. % of the quartz-rich regions (of which about half now comprises feldspar) and also formed much larger, scattered pockets (Fig. 10). During cooling, overgrowths formed on the quartz restite until the liquid moved onto the quartz–feldspar cotectic. At this point feldspar nucleated, and quartz and feldspar grew simultaneously to form granophyric intergrowths in the largest pores. CL images of the largest granophyre-filled pores demonstrate that undulose quartz–feldspar boundaries signify simultaneous crystallization of quartz and feldspar. In contrast, the smooth quartz–feldspar grain boundaries in the smaller pores denote sequential growth, with feldspar (or clinopyroxene or Fe-oxides in other parts of the rock; Fig. 8a) crystallizing alone, following quartz overgrowth on the pore walls. The critical pore width at which this change occurs is 250 µm (Fig. 8c). It should be noted that in some pores quartz continued crystallizing until the pore was completely filled (Fig. 10).
There is no kinetic barrier to overgrowth of pore walls, but crystallization of the phases not forming the pore walls requires homogeneous nucleation. Although this becomes easier as overgrowth on the walls proceeds (which increases the concentration of the other phases in the remaining liquid), the smaller the pore, the greater the supersaturation required for nucleation. Nucleation of the minor phase therefore cannot occur until the temperature has decreased to create sufficient undercooling to overcome the barrier to nucleation. Once nucleation does occur, the small pores are pseudomorphed by a few grains, which spread out in the porosity over 1–2 restitic grain diameters. In the case of the Biwabik migmatite, this plagioclase growth must have been accompanied by quartz overgrowth elsewhere, as the rejected quartz component migrated through the remaining porosity away from the growing feldspar.
Pore-size-controlled nucleation delay of the plagioclase in the Biwabik sample, ELY-6B, is consistent with the general increase in Na content of plagioclase in progressively smaller pores (Fig. 11). Within any single pore, the variation in Na content shows that growth occurred first in the widest region and subsequently propagated into the smallest extremities (Fig. 11c). When all the data are taken as a whole it appears that significantly delayed nucleation occurs in melt films below about 40 µm thickness. For ELY-6B, reduced solidification rates associated with delayed feldspar nucleation in the small pores permitted attainment of smoothly curved pore walls and (perhaps) also the equilibrium melt–quartz–quartz dihedral angle (20°, Holness, 2006). The change from Ab50 to Ab90 corresponds to a solidus temperature decrease of 30°C in the Qtz–Ab–An–H2O system at 5 kbar (Johannes, 1978). Although the cooling rate of the aureole of the Duluth Complex is not known, it is likely that a decrease in temperature of several tens of degrees would take longer than the few hundreds of years thought to be required to equilibrate a melt-bearing rock with a 1 mm grain size (Cheadle, 1989; Holness & Siklos, 2000).
A critical issue to address is the preservation of melt–solid dihedral angles in pseudomorphed pores. In the Biwabik rock, nucleation delay appears to have promoted melt–solid textural equilibration, and it is these equilibrium angles that are retained by the plagioclase pore-filling. However, in the Ashuanipi granulite, and perhaps also in the Opatica migmatite, dihedral angles recorded by the K-feldspar pseudomorphs are far from textural equilibrium. This implies both that any nucleation delay was insufficient to promote melt–solid textural equilibration before the melt was replaced by K-feldspar, and also that overgrowth on the walls of K-feldspar-filled pores in the vicinity of the pore corners was minimal. To interpret the preserved angle population in terms of melt distribution and subsequent cooling history it is essential to be certain that this population accurately reflects melt–solid angles, rather than being an artefact of overgrowth on the pore walls. Although a mechanism for pseudomorphing based on the kinetics of nucleation in small pores can be used to explain the observed microstructure in the Biwabik migmatites, there is no corresponding information on the progressive stages of solidification of either the Ashuanipi or the Opatica migmatites.
The Opatica sample described here contains a pseudomorph population that can plausibly account for a granitic liquid (i.e. quartz + plagioclase + microcline). All three phases crystallized together in larger pools and the thicker melt rims, whereas isolated single grains infill the smaller pores and thinner melt rims. Restitic plagioclase grains commonly have either overgrowths or elongate extensions into adjacent inferred melt pools, discernible by a variation in extinction position (Fig. 15a). In contrast, the Ashuanipi granulite is dominated by K-feldspar pseudomorphs. The absence of preserved CL information means that we cannot be sure about the quartz pore walls, but plagioclase pore walls bounding K-feldspar pseudomorphs appear optically to be compositionally uniform, suggesting that at least the plagioclase has not been overgrown. It is not clear why some melt-filled pores in the Ashuanipi migmatite were apparently perfectly pseudomorphed by K-feldspar, with the plagioclase and quartz components of the melt crystallizing elsewhere.
|
INTERPRETATION OF PSEUDOMORPHED PORE SHAPES |
|---|
Melt–solid textural equilibrium is most likely to have been attained in the Biwabik sample, ELY-6B, which contained about 20 vol. % melt within the quartz-rich regions (Fig. 10). This is close to both the disaggregation porosity of 23% and the minimum energy porosity of 18·5% for a dihedral angle of 20° (Cheadle, 1989
). Distinction between the possibilities that the quartz-rich parts of the rock formed a crystal slurry at the metamorphic peak, or that their microstructure records complete textural equilibrium, with excess melt expelled to form large pockets (such as that shown in Fig. 10; Jurewicz & Watson, 1985) is difficult, as the melt percentage is variable on the millimetre scale.
If textural equilibrium were achieved in rocks containing 5 vol. % melt, inferred for the Ashuanipi and Opatica samples, melt–solid dihedral angles of 20° would lead to most of the melt residing in tubular channels on three-grain junctions (Smith, 1964; Beeré, 1975). A randomly oriented population of such junctions will be predominantly sectioned at high angles to their length, so tubular melt pockets would generally appear as cuspate pockets at three-grain junctions. The observed dominance of elongate pockets on two-grain junctions in both EL180 and DL96-1006A cannot be accounted for by random cross-sectioning (see Wark et al., 2003) but must reflect a dominance of grain-boundary melt films, suggesting textural disequilibrium on the grain scale (even if equilibrium dihedral angles are established on a smaller scale). Melt films are most common on grain boundaries between reactant phases, suggesting that the grain-scale melt distribution is controlled predominantly by reaction kinetics.
Reported examples of dihedral angles associated with pore pseudomorphs (Holness & Clemens, 1999; Rosenberg & Riller, 2000; Marchildon & Brown, 2002) only rarely form populations consistent with either melt–solid (e.g. Rosenberg & Riller, 2000) or solid–solid textural equilibrium. This is in agreement with the results presented here for the contact metamorphosed rock from the Biwabik Formation, in which the unimodal plagioclase–quartz–quartz dihedral angle population (Fig. 9b) is intermediate between melt–solid and solid–solid textural equilibrium (Fig. 9a). A comparison with partially re-equilibrated angle populations in olivine cumulates (Holness et al., 2005b), in which a similar, intermediate, unimodal population occurs in clinopyroxene oikocrysts, suggests that the present dihedral angle population in the Biwabik Iron Formation sample represents the partial approach to solid-state textural equilibrium of an inherited fully equilibrated population of melt–solid dihedral angles, potentially providing a measure of the cooling rate in the sub-solidus.
On the assumption that the angles preserved in the Ashuanipi granulite reflect those inherited from the melt with no modification by pore-wall overgrowth, their strongly bimodal population is also clearly out of any kind of textural equilibrium, supportive of the wider disequilibrium recorded by the dominance of grain boundary reaction rims. As the films on plagioclase–quartz boundaries are commonly adjacent to reacting biotite grains, these films probably represent a propagating reaction front. The angle at propagating fronts is likely to be low and so melt–solid equilibration, at least in the immediate vicinity of the tip, would require relatively little adjustment. The absence of isolated K-feldspar lenses on plagioclase–quartz grain boundaries (see Fig. 1b) demonstrates that equilibration of the films, either melt–solid or solid–solid, did not occur.
The 90° peak, which represents the evolution of reaction-controlled geometries with an initial 180° angle, is below that expected for solid–solid textural equilibrium (with a median of 120°), suggesting that these angles were adjusting towards melt–solid equilibrium. The angle population comprises a significant proportion of angles between the two main peaks and these generally correspond to melt film geometries that appear to follow original grain boundaries, where the initial angle was likely to be in the range 120° < < 180° (Fig. 5).
Melt-filled pores pseudomorphed by K-feldspar in the Ashuanipi granulite thus appear to record a snapshot of melt geometries preserved at or close to the metamorphic peak, with only minor modifications in the sub-solidus. Melting occurred on grain boundaries between reacting phases, concentrated near biotite grains and propagating out along plagioclase–quartz grain boundaries. Reaction occurred sufficiently fast that equilibration of the microstructures was not possible apart from perhaps establishment of the equilibrium dihedral angle in the immediate vicinity of pore corners that already had low angles.
The geometry of the pseudomorphed melt in the Opatica amphibolite-facies migmatite is very different from that in both the Biwabik chert and the Ashuanipi granulite. Although thin cuspate interserts are common, much of the melt has been replaced by poly-mineralic aggregates on grain boundaries (Fig. 15a and b). These are highly reminiscent of the string-of-beads texture shown in Fig. 6d, and we suggest that their coarser grain size and more equilibrated texture (i.e. straight grain boundaries and 120° triple junctions) are due to annealing and equilibration in the sub-solidus. The melt films in the Opatica migmatite that solidified as poly-crystalline aggregates are significantly thicker than those replaced by microcline alone (Fig. 15a and b), reflecting the greater ease of nucleation within the larger pores.
The population of angles preserved at the corners of microcline-filled melt pores in the Opatica amphibolite-facies migmatite shows the same peak at 90° seen in the Ashuanipi sample, but the lower peak has a median of 60°, instead of 20°. The large driving force for grain boundary migration near the tips of pseudomorphed reaction-generated melt films permitted substantial approach to solid–solid textural equilibrium, whereas the inherited angles at the terminations of intra-phase boundaries at the walls of thick melt films did not evolve as much, because of their generally higher angles.
The relatively uncommon, isolated, pseudomorphs filled by quartz and plagioclase have very different shapes from those filled by microcline, with angles close to, or in sub-solidus textural equilibrium (Fig. 15b). This is probably because the adjacent grains are quartz and plagioclase and so microstructural adjustment towards lower energy configurations requires only mass movement across the grain boundary (Hunter, 1987). For the microcline (and the K-feldspar in the Ashuanipi granulite) there is no corresponding K-feldspar in the pore walls and so sub-solidus textural equilibration is much slower. The important corollary of this is that recrystallization and sub-solidus equilibration of inherited melt microstructures is dependent on the minerals forming the pseudomorphs. They are likely to retain a shape that is recognizable as a pseudomorph of a melt-filled pore only if they are formed of the minor component.
The three samples thus show a range of histories recorded by the geometry of the melt-filled pores. The Biwabik rock is likely to have attained complete melt–solid textural equilibrium, not only with equilibrated melt–solid dihedral angles but also possibly with an equilibrated pore geometry on the grain scale. This was followed by a partial approach to sub-solidus textural equilibrium during cooling. The Ashuanipi granulite records an almost unmodified reaction texture in which only partial approach to melt–solid equilibrium occurred, and the Opatica amphibolite records a reaction texture that has been significantly modified in the sub-solidus.
It is somewhat surprising that textural equilibrium was not attained, even on the scale of the pore corners, while melt was present in the granulite-facies regional (Ashuanipi) migmatite, whereas the Biwabik sample demonstrates the attainment of melt–solid textural equilibrium during contact metamorphism. The contrast may be due to the slightly higher temperatures in the Biwabik sample (>900°C compared with 820–900°C) or perhaps to the relative mineralogical simplicity of the Biwabik migmatite. Additionally, the quartz grains in the Biwabik rock record a complex thermal history and this may have contributed to a prolonged period of (fairly) constant high temperature during which reaction was slowed and melt–solid equilibration was facilitated.
The two regional terranes studied here record very different extents of sub-solidus textural equilibrium. This difference may have been a function of thermal history, with the Ashuanipi granulite cooling much more rapidly than the Opatica rocks as a consequence of their very different tectonic settings. This is consistent with the presence of microcline in the latter, and untwinned K-feldspar in the former. The Opatica Subprovince is thought to be the deep roots of a mountain chain whereas the Ashuanipi Subprovince is an ancient accretionary prism. It is highly likely that the Ashuanipi samples experienced rapid exhumation as a result of tectonism and active faulting in a terrane that remained at, or close to, a cratonic margin, whereas the Opatica rocks could have been exhumed only by erosion.
However, a further possibility is that the difference in sub-solidus textural equilibration was caused by the difference in melting reactions (with the Ashuanipi terrane undergoing dehydration melting whereas melting of the Opatica rocks was probably fluxed by externally derived H2O). The presence of H2O is known to facilitate textural equilibration in the sub-solidus (Holness et al., 1991) and it is possible that sufficient free H2O was present after melting in the Opatica rocks, whereas the grain boundaries in the Ashaunipi rocks remained dry. A much slower rate of sub-solidus textural maturation for the Ashuanipi rocks would avoid the necessity for an abrupt decrease in temperature coinciding with the almost complete depletion of K-feldspar that is required to preserve reaction textures in a residual gneiss. Given the potential of microstructural analysis for the decoding of migmatite history it is important that this question is addressed in future work.
|
SUMMARY AND CONCLUSIONS |
|---|
Comparison of reported microstructures in migmatites from a variety of crustal environments shows that the time scale of metamorphism exerts a crucial control on melt distribution and on the microstructures formed during solidification. Pseudomorphing of pores to form cuspate grains and pockets on grain boundaries generally necessitates temperature–time histories sufficient to permit melt migration. This is because pore pseudomorphing is dependent on there being a barrier to nucleation—such a barrier occurs in a small pore.
Observation of the distribution and shape of pseudomorphed pores in migmatites from a range of crustal environments demonstrates that the balance between textural equilibration, melting and deformation is only rarely in favour of textural equilibrium in melt-bearing rocks, even in regionally metamorphosed granulites. Observed low dihedral angles only rarely form populations consistent with melt–solid equilibrium (e.g. Rosenberg & Riller, 2000), and even where melt–solid equilibrium is achieved at pore corners, the grain-scale distribution may still be dominated by grain boundary melt films rather than channels on three-grain junctions.
Because the smallest pores are those that require the greatest supersaturation, these will be the last to solidify. The minerals filling the smallest pores will thus have more evolved compositions than grains of the same mineral in larger pores. A further consequence of the later in-filling of the smallest pores is that these will show the lowest solid-state dihedral angles. This is because they will have had a shorter time in which to undergo solid-state textural maturation compared with the earlier-formed interstitial grains in the larger pores (see Holness et al., 2005b). The control of pore size on solidification kinetics means that, given the same rate of cooling, a migmatite in which the melt is distributed in many small pores will take longer to solidify than one in which the same amount of melt is contained in fewer, larger, pores. This will affect the rheological response to cooling.
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to Stephen Reed for assistance with the CL and for suggesting that we examine Ti in quartz. Chris Hayward is thanked for help with microprobe analyses. Helpful comments from Jon Blundy, Jean-Louis Vigneresse, Richard White and Gary Stevens improved and clarified the manuscript.
*Corresponding author. E-mail: marian{at}esc.cam.ac.uk
|
REFERENCES |
|---|
Acosta-Vigil A, London D, Morgan G. B. VI. Experiments on the kinetics of partial melting of a leucogranite at 200 MPa H2O and 690–800°C: compositional variability of melts during the onset of H2O-saturated crustal anatexis. Contributions to Mineralogy and Petrology (2006) 151:539–557.[CrossRef][Web of Science]
Adamson AW. Physical Chemistry of Surfaces (1990) 5th. New York: John Wiley.
Ave’Lallemant HG, Carter NL. Syntectonic recrystallisation of olivine and modes of flow in the upper mantle. Geological Society of America Bulletin (1970) 81:2203–2220.
Beeré W. A unifying theory of the stability of penetrating liquid phases and sintering pores. Acta Metallurgica (1975) 23:131–138.[CrossRef][Web of Science]
Bigg EK. The supercooling of water. Proceedings of the Physical Society (London) (1953) 66B:688–694.[CrossRef]
Brown M. Orogeny, migmatites and leucogranites: a review. Proceedings of the Indian Academy of Science (Earth and Planetary Science) (2001) 110:313–336.
Brown M. Crustal melting and melt extraction, ascent and emplacement in orogens: mechanisms and consequences. Journal of the Geological Society, London (2007) 164:709–730.
Büsch W, Schneider G, Mehnert KR. Initial melting at grain boundaries. Part II: melting in rocks of granodioritic, quartzdioritic and tonalitic composition. Neues Jahrbuch für Mineralogie, Abhandlungen (1974) 1974:345–370.
Butler BCM. Metamorphism and metasomatism of rocks of the Moine Series by a dolerite plug in Glenmore, Ardnamurchan. Mineralogical Magazine (1961) 32:866–897.
Cahn JW. Surface stress and the chemical equilibrium of small crystals: I. The case of the isotropic surface. Acta Metallurgica (1980) 28:1333–1338.[CrossRef][Web of Science]
Cahn JW, Hoffman DW. A vector thermodynamics for anisotropic surfaces, II. Application to curved surfaces. Acta Metallurgica (1974) 22:1205–1214.[CrossRef][Web of Science]
Card KD, Ciesielski A. DNAG#1. Subdivisions of the Superior Province of the Canadian Shield. Geoscience Canada (1986) 13:5–13.[Web of Science]
Cesare B. Incongruent melting of biotite to spinel in a quartz-free restite at El Joyazo (SE Spain): textures and reaction characterisation. Contributions to Mineralogy and Petrology (2000) 139:273–284.[CrossRef][Web of Science]
Cheadle MJ. Properties of texturally equilibrated two-phase aggregates. In: Ph.D. thesis (1989) University of Cambridge. 166.
Clemens JD, Holness MB. Textural evolution and partial melting of arkose in a contact aureole: a case study and implications. Electronic Geosciences (2000) 5:4.[Medline]
Clemens JD, Mawer CK. Granitic magma transport by fracture propagation. Tectonophysics (1992) 204:339–360.[CrossRef][Web of Science]
Cmíral M, Fitz Gerald JD, Faul UH, Green DH. A close look at dihedral angles and melt geometry in olivine-basalt aggregates; a TEM study. Contributions to Mineralogy and Petrology (1997) 130:336–345.[Web of Science]
Connolly J, Holness M, Rubie D, Rushmer T. Reaction-induced microcracking: an experimental investigation of a mechanism for enhancing anatectic melt extraction. Geology (1997) 25:591–594.
Daines MJ, Kohlstedt DL. A laboratory study of melt migration. Philosophical Transactions of the Royal Society of London, Series A (1993) 342:43–52.[CrossRef]
Daines MJ, Kohlstedt DL. Influence of deformation on melt topology in peridotites. Journal of Geophysical Research (1997) 102:10257–10272.[CrossRef]
Elliott MT, Cheadle MJ, Jerram DA. On the identification of textural equilibrium in rocks using dihedral angle measurements. Geology (1997) 25:355–358.
Grant JA, Frost BR. Contact metamorphism and partial melting of pelitic rocks in the aureole of the Laramie anorthosite complex, Morton Pass, Wyoming. American Journal of Science (1990) 290:425–472.
Guernina S, Sawyer EW. Large-scale melt-depletion in granulite terranes: an example from the Archean Ashuanipi Subprovince of Quebec. Journal of Metamorphic Geology (2003) 21:181–201.[CrossRef][Web of Science]
Hammouda T, Laporte D. Ultrafast mantle impregnation by carbonatite melts. Geology (2000) 28:283–285.
Harte B, Hunter RH, Kinny PD. Melt geometry, movement and crystallisation, in relation to mantle dykes, veins and metasomatism. Philosophical Transactions of the Royal Society of London, Series A (1993) 342:1–21.[CrossRef]
Harte B, Pattison DRM, Linklater CM. Field relations and petrography of partially melted pelitic and semi-pelitic rocks. In: Equilibrium and Kinetics in Contact Metamorphism: the Ballachulish Igneous Complex and its Aureole—Voll G, Töpel J, Pattison DRM, Seifert F, eds. (1991) Berlin: Springer. 181–210.
Herring C. Surface tension as a motivation for sintering. In: Physics of Powder Metallurgy—Kingston WE, ed. (1951) New York: McGraw–Hill. 143–179.
Higgins MD. Quantitative Textural Measurements in Igneous and Metamorphic Petrology (2006) Cambridge: Cambridge University Press. 265.
Hiraga T, Nishikawa O, Nagase T, Akizuki M, Kohlstedt DL. Interfacial energies for quartz and albite in pelitic schist. Contributions to Mineralogy and Petrology (2002) 143:664–672.[Web of Science]
Hirth JG, Kohlstedt DL. Experimental constraints on the dynamics of the partially molten upper mangle: deformation in the diffusion creep regime. Journal of Geophysical Research (1995) 100:1981–2001.[CrossRef]
Hoffman DW, Cahn JW. A vector thermodynamics for anisotropic surfaces, I. Fundamentals and applications to plane surface junctions. Surface Science (1972) 31:368–388.[CrossRef][Web of Science]
Holness MB. Equilibrium dihedral angles in the system quartz–CO2–H2O–NaCl at 800°C and 1–15 kbar: the effects of pressure and fluid composition on the permeability of quartzites. Earth and Planetary Science Letters (1992) 114:171–184.[CrossRef][Web of Science]
Holness MB. Contact metamorphism and anatexis of Torridonian arkose by minor intrusions of the Rum Igneous Complex, Inner Hebrides, Scotland. Geological Magazine (1999) 136:527–542.[Abstract]
Holness MB. Growth and albitisation of K-feldspar in crystalline rocks in the shallow crust: a tracer for fluid circulation? Geofluids (2003) 3:89–102.[CrossRef][Web of Science]
Holness MB. Melt–solid dihedral angles of common minerals in natural rocks. Journal of Petrology (2006) 47:791–800.
Holness MB, Clemens JD. Partial melting of the Appin Quartzite driven by fracture-controlled H2O infiltration in the aureole of the Ballachulish Igneous Complex, Scottish Highlands. Contributions to Mineralogy and Petrology (1999) 136:154–168.[CrossRef][Web of Science]
Holness MB, Isherwood C. The aureole of the Rum Tertiary Igneous Complex, Inner Hebrides, Scotland. Journal of the Geological Society, London (2003) 160:15–27.
Holness MB, Siklos STC. Rates of textural equilibration in fluid-bearing systems: kinetic limitations to surface-energy controlled permeability. Chemical Geology (2000) 162:137–153.[CrossRef][Web of Science]
Holness MB, Watt GR. The aureole of the Traigh Bhan na Sgurra sill, Isle of Mull: reaction-driven micro-cracking during pyrometamorphism. Journal of Petrology (2002) 43:511–534.
Holness MB, Bickle MJ, Graham CM. On the kinetics of textural equilibration in forsterite marbles. Contributions to Mineralogy and Petrology (1991) 108:356–367.[CrossRef][Web of Science]
Holness MB, Dane K, Sides R, Richardson C, Caddick M. Melting and melt segregation in the aureole of the Glenmore Plug, Ardnmurchan. Journal of Metamorphic Geology (2005a) 23:29–43.[CrossRef][Web of Science]
Holness MB, Cheadle MJ, McKenzie D. On the uses of changes in dihedral angle to decode late-stage textural evolution in cumulates. Journal of Petrology (2005b) 46:1565–1583.
Holness MB, Anderson AT, Martin VM, Maclennan J, Passmore E, Schwindinger K. Textures in partially solidified crystalline nodules: a window into the pore structure of slowly cooled mafic intrusions. Journal of Petrology (2007) 48:1243–1264.
Hunter RH. Textural equilibrium in layered igneous rocks. In: Origins of Igneous Layering—Parsons I, ed. (1987) Dordrecht: D. Reidel. 473–503.
Jin Z, Green HW, Zhou Y. Melt topology in partially molten mantle peridotite during ductile deformation. Nature (1994) 372:164–167.[CrossRef]
Johannes W. Melting of plagioclase in the system Ab–An–H2O and Qz–Ab–An–H2O at PH2O = 5 kbars, an equilibrium problem. Contributions to Mineralogy and Petrology (1978) 66:295–303.[CrossRef][Web of Science]
Jurewicz SR, Jurewicz AJG. Distribution of apparent angles on random sections with emphasis on dihedral angle measurements. Journal of Geophysical Research (1986) 91:9277–9282.
Jurewicz SR, Watson EB. The distribution of partial melt in a granitic system: the application of liquid phase sintering theory. Geochimica et Cosmochimica Acta (1985) 49:1109–1121.[CrossRef][Web of Science]
Kenah P, Hollister LS. Anatexis in the Central Gneiss complex, British Columbia. In: Migmatites, Melting and Metamorphism—Atherton MP, Gribble CD, eds. (1983) Nantwich: Shiva. 142–162.
Kohlstedt DL, Zimmerman ME. Rheology of partially molten mantle rocks. Annual Review of Earth and Planetary Sciences (1996) 24:41–62.[CrossRef][Web of Science]
Kretz RH. Interpretation of the shape of mineral grains in metamorphic rocks. Journal of Petrology (1966) 7:68–94.
Laporte D, Provost A. Equilibrium geometry of a fluid phase in a polycrystalline aggregate with anisotropic surface energies: Dry grain boundaries. Journal of Geophysical Research (2000) 105:25937–25953.[CrossRef]
Laporte D, Watson EB. Experimental and theoretical constraints on melt distribution in crustal sources: the effect of crystalline anisotropy on melt interconnectivity. Chemical Geology (1995) 124:161–184.[CrossRef][Web of Science]
Laporte D, Rapaille C, Provost A. Wetting angles, equilibrium melt geometry, and the permeability threshold of partially molten crustal protoliths. In: Granite: From Segregation of Melt to Emplacement Fabrics—Bouchez J.-L, Hutton DH, Stephens WE, eds. (1997) Norwell, MA: Kluwer Academic. 31–54.
Lupulescu A, Watson EB. Low melt fraction connectivity of granitic and tonalitic melts in a mafic crustal rock at 800°C and 1 GPa. Contributions to Mineralogy and Petrology (1999) 134:202–216.[CrossRef][Web of Science]
Maddock RH. Partial melting of lithic porphyroclasts in fault-generated pseudotachylytes. Neues Jahrbuch für Mineralogie, Abhandlungen (1986) 155:1–14.[Web of Science]
Marchildon N, Brown M. Grain-scale melt distribution in two contact aureole rocks: implications for controls on melt localisation and deformation. Journal of Metamorphic Geology (2002) 20:381–396.[CrossRef][Web of Science]
Maumus J, Laporte D, Schiano P. Dihedral angle measurements and infiltration property of SiO2-rich melts in mantle peridotite assemblages. Contributions to Mineralogy and Petrology (2004) 148:1–12.[Web of Science]
Mehnert KR, Büsch W, Schneider G. Initial melting at grain boundaries of quartz and feldspar in gneisses and granulites. Neues Jahrbuch für Mineralogie, Monatshefte (1973) 4:165–183.
Melia TP, Moffitt WP. Crystallisation from aqueous solution. Journal of Colloid Science (1964) 19:433–447.[CrossRef][Web of Science]
Morey GB. Chemical composition of the Eastern Biwabik Iron formation (Early Proterozoic), Mesabi Range, Minnesota. Economic Geology (1992) 87:1649–1658.
Paquet J, François P. Experimental deformation of partially melted granitic rocks at 600–900°C and 250 MPa confining pressure. Tectonophysics (1980) 68:131–146.[CrossRef][Web of Science]
Park HH, Yoon DN. Effect of dihedral angle on the morphology of grains in a matrix phase. Metallurgical Tranactions (1985) 16:923–928.[CrossRef]
Pattison DRM, Harte B. Evolution of structurally contrasting anatectic migmatites in the 3-kbar Balachulish aureole, Scotland. Journal of Metamorphic Geology (1988) 6:475–494.[Web of Science]
Percival JA. A regional perspective of the Quetico meta-sedimentary belt, Superior Province, Canada. Canadian Journal of Earth Sciences (1989) 26:677–693.
Percival JA. Granulite-facies metamorphism and crustal magmatism in the Ashuanipi complex, Quebec–Labrador, Canada. Journal of Petrology (1991) 32:1261–1297.
Percival JA, Mortensen JK, Stern RA, Card K, Begin NJ. Giant granulite terranes of northeastern Superior Province: the Ashuanipi complex and Minto block. Canadian Journal of Earth Sciences (1992) 29:2287–2308.
Platten IM. Partial melting of feldspathic quartzite around late Caledonian minor intrusions in Appin, Scotland. Geological Magazine (1982) 119:413–419.[Abstract]
Platten IM. Partial melting of semi-pelite and the development of marginal breccias around late Caledonian minor intrusives in the Grampian Highlands of Scotland. Geological Magazine (1983) 120:37–49.[Abstract]
Price JD, Wark DA, Watson EB, Smith AM. Grain-scale permeabilities of faceted polycrystalline aggregates. Geofluids (2006) 6:302–318.[CrossRef][Web of Science]
Putnis A, Mauthe G. The effect of pore size on cementation in porous rocks. Geofluids (2001) 1:37–41.[CrossRef]
Riller U, Cruden AR, Schwerdtner WM. Magnetic fabric and microstructural evidence for a tectono-thermal overprint of the early Proterozoic Murray pluton, central Ontario, Canada. Journal of Structural Geology (1996) 18:1005–1016.[CrossRef][Web of Science]
Rosenberg CL, Riller U. Partial melt topology in statically and dynamically recrystallised granite. Geology (2000) 28:7–10.
Rushmer T. Volume change during partial melting reactions: implications for melt extraction, melt geochemistry and crustal rheology. Tectonophysics (2001) 342:389–405.[CrossRef][Web of Science]
Sawyer EW. Formation and evolution of granite magmas during crustal reworking: the significance of diatexites. Journal of Petrology (1998) 39:1147–1167.[CrossRef][Web of Science]
Sawyer EW. Criteria for the recognition of partial melting. Physics and Chemistry of the Earth (1999) 24:269–279.[CrossRef][Web of Science]
Sawyer EW. Melt segregation in the continental crust: distribution and movement of melt in anatectic rocks. Journal of Metamorphic Geology (2001) 19:291–309.[CrossRef][Web of Science]
Sawyer EW, Benn K. Structure of the high-grade Opatica Belt and adjacent low-grade Abitibi subprovince, Canada: an Archaean mountain front. Journal of Structural Geology (1993) 15:1443–1458.[CrossRef][Web of Science]
Scherer GW. Crystallisation in pores. Cement and Concrete Research (1999) 29:1347–1358.[CrossRef][Web of Science]
Smith CS. Grains, phases and interfaces: an interpretation of microstructure. Transactions of the Metallurgical Society of the AIME (1948) 175:15–51.
Smith CS. Some elementary principles of polycrystalline microstructure. Metallurgical Reviews (1964) 9:1–48.
Urai JL. Water assisted dynamic recrystallisation and weakening in polycrystalline bischoffite. Tectonophysics (1983) 96:125–157.[CrossRef][Web of Science]
Van Schmus WR, Hinze WJ. The Midcontinent rift. Annual Review of Earth and Planetary Sciences (1985) 13:345–383.[CrossRef][Web of Science]
Vernon RH. Microstructures of high-grade metamorphic rocks at Broken Hill, Australia. Journal of Petrology (1968) 9:1–22.
Vernon RH. Comparative grain-boundary studies of some basic and ultrabasic granulites, nodules and cumulates. Scottish Journal of Geology (1970) 6:337–351.[CrossRef]
Vernon RH. On the identification of textural disequilibrium in rocks using dihedral angle measurements: Comment. Geology (1997) 25:1055–1055.
Vernon RH. A Practical Guide to Rock Microstructure (2004) Cambridge: Cambridge University Press. 594.
Vernon RH, Collins WJ. Igneous microstructures in migmatites. Geology (1988) 16:1126–1129.
Von Bargen N, Waff HS. Permeability, interfacial areas and curvatures of partially molten systems: results of numerical computations of equilibrium microstructures. Journal of Geophysical Research (1986) 91:9261–9276.
Walte NP, Bons PD, Passchier CW. Deformation of melt-bearing systems—insight from in situ grain-scale analogue experiments. Journal of Structural Geology (2005) 27:1666–1679.[CrossRef][Web of Science]
Wark DA, Spear FS. Titanium in quartz: cathodoluminscence and thermometry. Geochimica et Cosmochimica Acta (2005) 69:A592.
Wark DA, Watson EB. Grain-scale permeabilities of texturally equilibrated, monomineralic rocks. Earth and Planetary Science Letters (1998) 164:591–605.[CrossRef][Web of Science]
Wark DA, Watson EB. The TitaniQ: a titanium-in-quartz geothermometer. Contributions to Mineralogy and Petrology (2006) 152:743–754.[CrossRef][Web of Science]
Wark DA, Williams CA, Watson EB, Price JD. Reassessment of pore shapes in microstructurally equilibrated rocks, with implications for permeability of the upper mantle. Journal of Geophysical Research (2003) 108. doi:10.1029/2001JB001575.
Watson EB. Melt infiltration and magma evolution. Geology (1982) 10:236–240.
Watson EB, Brenan JM. Fluids in the lithosphere, I: Experimentally-determined wetting characteristics of CO2–H2O fluids and their implications for fluid transport, host rock physical properties, and fluid inclusion formation. Earth and Planetary Science Letters (1987) 85:497–515.[CrossRef][Web of Science]
Wiebe RA, Wark DA, Hawkins DP. Insights from quartz cathodoluminescence zoning into crystallisation of the Vinalhaven granite, coastal Maine. Contributions to Mineralogy and Petrology (2007) 154:439–453.[CrossRef][Web of Science]
Yoshino T, Price JD, Wark DA, Watson EB. Effect of faceting on pore geometry in texturally equilibrated rocks: implications for low permeability at low porosity. Contributions to Mineralogy and Petrology (2006) 152:169–186.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Cesare, S. Ferrero, E. Salvioli-Mariani, D. Pedron, and A. Cavallo "Nanogranite" and glassy inclusions: The anatectic melt in migmatites and granulites Geology, July 1, 2009; 37(7): 627 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Plumper and A. Putnis The Complex Hydrothermal History of Granitic Rocks: Multiple Feldspar Replacement Reactions under Subsolidus Conditions J. Petrology, May 18, 2009; (2009) egp028v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. R. NICOLL, M. B. HOLNESS, V. R. TROLL, C. H. DONALDSON, E. P. HOLOHAN, C. H. EMELEUS, and D. CHEW Early mafic magmatism and crustal anatexis on the Isle of Rum: evidence from the Am Mam intrusion breccia Geological Magazine, May 1, 2009; 146(3): 368 - 381. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||